April 26, 2010 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
April 26, 2010 [00:00:00] <CosmicDJ> you can enable smb sharing with zfs[00:00:15] <vraa> i think that's what i might've done on accident[00:00:17] *** ravenslay3r has left #opensolaris[00:00:26] <vraa> i think i enabled some share setting in the zpool properties?[00:00:46] <vraa> i'm trying to basically setup a way so certain users can't access certain folders[00:01:09] <vraa> i have about 10 users, 1 server, and perhaps 2 groups "management" "nonmanagement"[00:01:51] <datadigger> Hi. Situation: SXCE b130 installed on x86 laptop. Two pools in c0d0p1: rpool in c0d0s0, datapool in c0d0s1.[00:01:56] <datadigger> In my innocence I left 24 GB of free space behind c0d0p1 (unallocated), which I want to add to the datapool now.[00:02:03] <datadigger> I can run from an rpool mirror on external USB disk during the change (offline c0d0s0, export datapool), but I can't boot from USB, so it is a bit scary.[00:02:10] <datadigger> It appears I cannot resize c0d0p1. I would have to delete c0d0p1 and recreate it with fdisk, which would destroy the VTOC (says man -s 1M fdisk).[00:02:17] <datadigger> I don't mind to recreate the VTOC = redefine the slices with format (of course I made a note of all offsets and sizes).[00:02:24] <datadigger> Question: Will my zpools be intact? The next step would be to resize c0d0s1 to expand the datapool, which is easy.[00:03:04] <datadigger> TIA[00:03:32] *** dalibor_ has quit IRC[00:03:38] <tsoome> well. i wouldnt really recommend sharing same physical disk between 2 pools.[00:04:21] <vraa> tsoome: do you have any opinion of which is better, cifs in kernel, or samba with swat?[00:04:31] <nikolam> datadigger, I believe sxce will not be developed anymore. Care to switch on Opensolaris?[00:04:36] <tsoome> that depends a bit.[00:04:38] <datadigger> tsoome: True, but I like to have my data in a separate pool.[00:05:06] <datadigger> nikolam: I will switch to opensolaris some day. The separate datapool comes in handy then ;)[00:05:39] <tsoome> samba shares directory tree, not filesystem, and samba will allow to share printers as well. smb is faster as reported (as its implemented as kernel module)[00:05:46] <datadigger> (I think a fresh install of rpool is faster than a updating from sxce b130 to 2010.0x[00:06:27] <tsoome> as for sharing disk - if you have disk issue, both rpool and data pool are affected.[00:07:34] <vraa> tsoome thank you very much, so i cannot say "hey kernel cifs: share only vkc_zfs_tank/Home" it has to share vkc_zfs_tank completely[00:07:35] <datadigger> sure. That's where the USB rpool mirror is for. The datapool is backupped with zfs send | zfs recv[00:07:56] *** ineil has joined #opensolaris[00:07:57] <datadigger> s/where/what/[00:08:26] <datadigger> I'd love to have 4 disks in my laptop, but it doesn't fit.[00:08:31] <x58> vraa: You can share any ZFS file system, so if you have rpool/export/home/username you can use the in kernel CIFS to share it.[00:08:56] <tsoome> well, with only rpool on disk, you still can create separate dataset for data and use zfs send to back it up....[00:09:13] <tsoome> so you wont loose that and your setup is simpler.[00:09:33] <datadigger> tsoome: Thanks for the hint, I'll consider that.[00:11:13] *** dalibor has joined #opensolaris[00:11:34] <datadigger> Back to my question, will my pools be intact when I recreatie c0d0p1 with fdisk and redefine the slice(s) for the pool(s) ?[00:12:11] <tsoome> noone cant give quarantee, but most probably. anyhow, keep backup handy.[00:12:47] *** Spencer_tt has quit IRC[00:13:14] *** Spencer_tt has joined #opensolaris[00:14:01] *** piwi has quit IRC[00:16:03] <datadigger> tsoome: Thanks. Backup is taken care of on a mirrored backup pool on 2 USB disks. I'll try to get to boot the rpool mirror which I have on USB. Probably have to edit GRUBs findroot when I do that.[00:19:19] *** relling has joined #opensolaris[00:21:27] *** ineil has left #opensolaris[00:24:30] *** bahumbug has joined #opensolaris[00:26:20] *** niq has quit IRC[00:28:17] *** moepmoep has quit IRC[00:30:12] *** bahumbug has quit IRC[00:30:22] *** Odin- has quit IRC[00:30:26] *** bahumbug has joined #opensolaris[00:33:06] *** moepmoep has joined #opensolaris[00:35:02] *** sk0rd has quit IRC[00:36:12] *** ewdafa has quit IRC[00:36:27] <vraa> i have tried to start from scratch, i disabled the shares in sharemgr, and i disabled the shares in swat[00:36:50] <vraa> when i do[00:37:37] <vraa> "ps aux | grep smb" and kill all the things, there is still one "/usr/lib/smbsrv/sm" that keeps coming up[00:38:40] <vraa> i have done "svcadm disable samba" but it's still there[00:40:01] *** Seeder has quit IRC[00:40:38] <vraa> i will study more and try again later[00:40:42] *** vraa has quit IRC[00:41:36] <tsoome> as i wrote, in kernel smb has nothing to do with samba; therefore svcadm disable samba wont have any effect and /usr/lib/smbsrv/sm is not managed by it;)[00:43:52] *** dalibor has quit IRC[00:48:21] *** jpd1234 has quit IRC[00:49:05] *** bahumbug has quit IRC[00:51:17] *** nanase has quit IRC[00:52:03] *** sndcrb has quit IRC[00:52:57] *** bahumbug has joined #opensolaris[00:53:18] *** Odin- has joined #opensolaris[00:58:17] *** bahumbug has quit IRC[01:03:23] *** RoyK has joined #opensolaris[01:15:07] *** Terlisimo has joined #opensolaris[01:15:09] *** Esteban25 has joined #opensolaris[01:15:38] <Esteban25> hello[01:15:51] <Esteban25> can anyone help me out here with somtthing?[01:16:23] <Esteban25> is there something like apt-get for open solaris 2009?[01:16:32] <RoyK> pkg[01:16:41] <Esteban25> pkg install?[01:17:17] <tsoome> man pkg[01:17:21] <Esteban25> lol[01:17:22] <Esteban25> sorry[01:17:31] <Esteban25> i was doing pkg-get[01:17:45] <Esteban25> because i heard of pkg but i thought it was pkg-get[01:18:50] *** Mdx4 has quit IRC[01:18:55] *** n3kl has quit IRC[01:23:51] *** boyd has quit IRC[01:25:50] *** RoyK has left #opensolaris[01:27:37] <Esteban25> damn[01:27:44] <Esteban25> pkg cant find memcached[01:29:41] *** psychicist has quit IRC[01:30:28] *** Terlisimo has quit IRC[01:30:49] <tsoome> http://blogs.sun.com/trond/entry/memcached_in_solaris is created 2 years ago....[01:30:52] *** Cobi has joined #opensolaris[01:31:16] *** sergiusens has joined #opensolaris[01:33:48] *** _madmax_ has joined #opensolaris[01:35:07] *** _madmax_ has quit IRC[01:35:24] *** _madmax_ has joined #opensolaris[01:38:04] *** Gman has joined #opensolaris[01:38:04] *** ChanServ sets mode: +o Gman[01:38:55] *** nevdull has joined #opensolaris[01:39:52] *** sergiusens has quit IRC[01:39:53] *** _madmax__ has joined #opensolaris[01:43:42] *** _madmax_ has quit IRC[01:43:43] *** _madmax__ is now known as _madmax_[01:44:44] *** achester has quit IRC[01:45:35] *** nevdull has quit IRC[01:46:25] * stallion grumbles.[01:46:31] <stallion> jmcp: alive?[01:46:51] <stallion> jbk: how about you?[01:49:43] <jbk> about to leave for practice[01:50:01] <stallion> quick question[01:50:26] <jbk> 42[01:50:28] <jbk> =][01:51:13] <stallion> Wondering what the damage would be updating the vmem directly rather than using ddi_put* for bits allocated by ddi_dma_mem_alloc[01:51:32] <stallion> more out of academic curiosity than anything else :D[01:51:43] <stallion> I would suspect a ddi_dma_sync to take care of any problems[01:52:11] <stallion> although I wonder if this is more of an issue on systems with dvma[01:53:09] <jbk> for memory mapped for dma, i think you should be able to read and write it directly (with suitable syncing)[01:53:20] <stallion> thats what I would have thought[01:53:25] <stallion> I even tried it with no ill effects[01:53:27] <jbk> i always thought the ddi_put was more for non dma stuff[01:53:42] <stallion> well, I looked at the man page, its for that purpose as well[01:54:07] *** achester has joined #opensolaris[01:54:14] <stallion> I've not found any rationale in WDD either[01:54:17] <jbk> i mean, the network drivers don't use ddi_put for copying mblk's[01:54:21] <stallion> I'm wondering if its more for free bswapping[01:54:24] *** icarus902 has joined #opensolaris[01:54:27] <jbk> that could be[01:54:41] <jbk> anyway.. gotta go..[01:54:47] <stallion> I'll bug garrett later ;)[01:54:47] <stallion> cya[01:55:15] <stallion> Ah![01:55:16] <stallion> Found it[01:55:43] <stallion> If DDI_NEVERSWAP_ACC is specified, memory can be used for any purpose; but if either endian mode is specified, you must use ddi_get/put* and never anything else.[01:56:06] <stallion> so I was right then - free bswapping ;)[01:57:00] <WebDawg> What is the disk naming scheme for O.S?[01:57:09] <stallion> c0t0d0s0 ?[01:57:15] <stallion> Channel TUN Disk Slice[01:57:21] <stallion> Controller rather[01:58:12] <WebDawg> I try and do zpool create blabla mirror c8d0 c8d1[01:58:28] <stallion> damnit[01:58:36] <WebDawg> when i do the format command[01:58:37] <stallion> I belive the 't' is for target[01:58:44] <stallion> I've had networking bits on my mind all day[01:58:46] <WebDawg> no t* is showing.[01:58:52] <stallion> its possible[01:58:58] <stallion> not all controllers support multiple targets[01:59:11] <WebDawg> How do I discover active disks?[01:59:15] <stallion> iostat -En[02:00:59] <WebDawg> http://pastebin.com/MucUDAr8[02:02:19] <WebDawg> wait[02:02:20] <WebDawg> http://pastebin.com/JfVQLgPH[02:04:19] <WebDawg> I guess I could add s0[02:04:31] <WebDawg> but why have a slice for something that isnt a slice.[02:05:00] <stallion> Is this for a boot device?[02:05:05] <WebDawg> no[02:05:20] <WebDawg> But I know zfs hates slices, or so they say.[02:05:33] <stallion> well[02:05:44] <stallion> using a slice impedes write performance[02:05:58] <stallion> did you label the disks?[02:06:03] *** Spencer_tt has quit IRC[02:06:08] <relling> ZFS only uses slices[02:06:16] <stallion> relling: what? no.[02:06:29] <WebDawg> lable the disks?[02:06:31] <WebDawg> nah[02:06:35] <stallion> zfs-boot requires the use of a slice, but zfs strongly prefers whole-disks[02:06:41] <stallion> WebDawg: issue: format -e[02:06:43] <joshua_> I never understood why slices / partitions impeded write performance -- why is that?[02:06:43] <WebDawg> should I dd the drives?[02:06:45] <relling> but if you give it a disk, it will label a slice which uses the whole disk and enable the write cache[02:06:50] <stallion> pick the target disks, and label it with EFI[02:07:26] <stallion> relling: there is no vtoc; don't re-use the term "slice"[02:07:29] <relling> then it turns around and hides "s0" from you, to add confusion into the mix[02:07:58] <WebDawg> stallion, it gave me the same device names.[02:08:11] <relling> EFI slices are presented as s0, s1, s2, ...[02:08:12] <WebDawg> except stuff like this stallion: /pci@0,0/pci-ide@1f,2/ide@1/cmdk@1,0[02:08:13] <stallion> WebDawg: devfsadm -C[02:08:21] <stallion> blech[02:08:30] <stallion> you need a better controller :P[02:08:44] <WebDawg> stallion, what did that do?[02:09:08] <stallion> WebDawg: it cleans up your /dev hierarchy - do you have entries for c8d0 etc. ?[02:09:11] <joshua_> oh, by the way -- does anyone know of a decent (read performance; don't care about reliability) PCIe SSD to use for L2ARC?[02:09:51] <joshua_> we want a good chunk of L2ARC so we can do dedup on a biggish (a few TB) dataset without too much pain[02:10:06] <relling> PCIe?[02:10:15] <WebDawg> http://pastebin.com/9kDHSv9J[02:10:22] <joshua_> relling, yes; we're out of SATA drive bays[02:10:46] <joshua_> (and also out of PATA drive bays, for that matter)[02:10:47] <relling> not many options for PCIe[02:11:05] <WebDawg> I still do not have JUST c8d0[02:11:09] *** agrajag has joined #opensolaris[02:12:22] *** Esteban25 has quit IRC[02:12:36] <relling> c8d0 does not exist[02:12:41] <joshua_> hmm, evidently these mini-pcie devices are mini-pcie in form factor, but in fact are SATA devices[02:13:18] <WebDawg> relling, why does format print it out:?[02:14:24] <relling> c8d0 is a reference to controller 8 disk 0. But that is a disk, not a partition or slice.[02:15:15] *** _madmax_ has quit IRC[02:16:30] *** freakazoid0223 has quit IRC[02:16:30] <stallion> WebDawg: so `zpool create tank mirror c8d0 c8d1` doesn't "work" for you ?[02:16:48] *** loke has quit IRC[02:21:11] *** _madmax_ has joined #opensolaris[02:25:32] *** eclark has joined #opensolaris[02:26:21] <eclark> i updated to the dev release, and now apache is segfaulting randomly[02:27:02] <eclark> 1.) is there somewhere i should report this, or check to see if it's already reported, and 2.) is there some way to revert just the apache package back to the non-dev version?[02:27:03] *** rmorse has quit IRC[02:28:22] <WebDawg> stallion, correct.[02:30:42] <WebDawg> stallion, this is a better post: http://pastebin.com/PYtrYwj8[02:32:14] <WebDawg> I am having this problem creating a zfs disk pool in opensolaris. I cannot understand how to do it without using SLICE 0 or s0 after the drives. From what everyone has been telling me I should use an entire disk. Please Help. http://pastebin.com/PYtrYwj8[02:37:48] *** freetown2 has joined #opensolaris[02:39:21] *** ma3x__ has joined #opensolaris[02:42:42] *** ma3x has quit IRC[02:43:12] *** ma3x__ is now known as ma3x[02:43:21] <WebDawg> Anyone?[02:44:09] <freetown2> ahci again?[02:46:23] <Andys^> WebDawg: interesting.[02:46:45] <Andys^> WebDawg: you might like to run the format utility and clear out any partitions or slices existing on the drive[02:51:12] <WebDawg> Andys^, can I redetect the stuff without restarting?[02:51:49] <Andys^> i think so..[02:52:03] *** andrnils has quit IRC[02:52:54] <WebDawg> the drives are written with random information atm[02:53:00] <WebDawg> no partitions on them at all.[02:54:41] <Andys^> well better to run format utility[02:54:54] <bdha> Shouldn't matter.[02:55:36] <bdha> e.g., http://southbrain.com/south/tutorials/zpools.html[02:55:39] <bdha> Should just work.[02:55:49] <bdha> Try running devfsadm -C.[02:56:15] <WebDawg> I figured it out. Its because I was not root.[02:56:19] <Andys^> "written with random information" sounds sus[02:56:27] <Andys^> are they brand new or used?[02:56:44] <bdha> WebDawg: Yeah.. that'll do it..[02:56:48] <WebDawg> used. they where truecrypt device volumes. They have never had any partitioning information on them.[02:57:20] <bdha> ZFS doesn't care what's on them.[02:57:26] <Andys^> it is actually possible the random data is confusing opensolaris label reader :)[02:57:37] <bdha> Andys^: Did you miss where he wasn't root? )[02:57:38] <bdha> :)[02:57:46] <WebDawg> They online now.[02:57:49] <WebDawg> bbl[02:57:53] <Andys^> no,m i assumed no one could be that stupid..[02:57:55] *** andrnils has joined #opensolaris[02:58:17] <WebDawg> one more thing. what is the ZFS for raid 0 no striping or parity?[02:58:37] <bdha> Nothing. You just say "create eatmydata disk1 disk2 disk3 .."[02:58:40] <Andys^> just use 'create'[02:58:41] * freetown2 grabs crisps and soda and settles down in flame proof bunker[02:58:54] <Andys^> zpool stripes raid0 across all vdevs, so you justcreate a pool where each disk is its own single-disk vdev[02:59:41] <WebDawg> k[03:00:02] <WebDawg> if I havnt written anything to the drive...[03:00:11] <WebDawg> Has it written anything to the drive?[03:00:36] <Andys^> during create? it would be labelling the disks and creating the databse yes..[03:05:15] *** andrnils has quit IRC[03:05:29] *** andrnils has joined #opensolaris[03:08:17] *** e1kg has joined #opensolaris[03:12:29] *** tosh has quit IRC[03:39:12] *** Norrin has quit IRC[03:40:43] *** Mace_N900 has joined #opensolaris[03:50:15] *** vraa has joined #opensolaris[03:50:35] <vraa> howdy guys, i am back, i finally figured out a way to disable the cifs kernel sharing AND the samba sharing, i am just trying to set things up through SWAT now[03:50:44] *** wdp_aao__ has joined #opensolaris[03:50:45] *** wdp__ has joined #opensolaris[03:51:12] <freetown2> vraa, so you are going to swat all the windows machines away?[03:51:21] <vraa> lol, i wish[03:51:29] <vraa> but even if i did i still wouldn't have setup samba ([03:51:41] <vraa> :( it's the osol box that i am having trouble with[03:52:12] <freetown2> oh...[03:52:38] <vraa> i got SWAT to "control" smb, now when i press START ALL or STOP ALL i can see all the /usr/sfw/sbin/smbd processes go away[03:53:10] <freetown2> nice...[03:53:27] <freetown2> svcadm not your thing?[03:54:07] *** wdp_aao_ has quit IRC[03:54:14] *** wdp_ has quit IRC[03:54:48] <vraa> i can't figure it out[03:55:06] <freetown2> ?[03:55:09] <vraa> BUT I THINK I GOT SWAT WORKING 100%! now when i browse the shares from a windows pc, the STATUS section updates properly and tell s me the IP and client and files used[03:55:23] <freetown2> smbstatus...nice...[03:55:49] <freetown2> maybe i should enable swat too...not used it in a while...[03:56:18] <freetown2> nah, it might choke on my ldap backend[03:56:20] <vraa> i downloaded many things called "open solaris bible" and i've been reading through them to learn "sharemgr" and "svcadm"[03:56:36] <vraa> ldap = active directory right? i tried to setup on on win2k8, that's what i'm trying to use for my user auth[03:57:27] <freetown2> vraa, well, no...i use winbind to sync with AD and use Sun DS as the backend for samba[03:58:57] <vraa> i thougth winbind was for hostname lookup[03:59:00] *** The-spiki has quit IRC[03:59:19] *** forquare has quit IRC[03:59:39] *** forquare has joined #opensolaris[03:59:51] <freetown2> huh? where did you get that idea from? winbind is for user uid/gid syncing between Unix/Linux systems and Windows[04:00:11] <freetown2> wins is for hostname lookup[04:00:51] <freetown2> mind you, they encourage using the kernel cifs server and not samba now i think[04:01:15] <freetown2> i cannot wait for samba 3.4.x to get added to OpenSolaris[04:02:07] <freetown2> cor...samba 3.5.x is out already?! i wonder what's available...[04:02:12] * freetown2 heads off[04:02:13] <vraa> i tried using kernel cifs, but i didn't see a different except it was more confusing to do sharemgr than SWAT gui[04:02:52] <freetown2> vraa, you mean the zfs man page is too hard for you?[04:03:18] <vraa> not too hard, i went through that and all the CIFS tasks list on the osol page, just need more experience to undrestand it properly[04:03:28] <vraa> in the mean time, SWAT works really nicely with /etc/sfw/smb.conf[04:04:24] <freetown2> extended the schema on windows did you?[04:04:48] <vraa> sorry, i do not understand the question[04:05:12] <freetown2> i thought that was something that needed doing to use the cifs kernel serer?[04:05:18] <freetown2> s/serer/server/[04:05:25] <vraa> right now i haven't integrated windows server 08 and osol, besides only thing i have on win2k8 are the following roles: active directory domain services (with 2 users) + dns server + dhcp server +win deployment services[04:05:43] <vraa> well i could get the cifs kernel working, using the zfs set sharesmb=on property[04:06:25] <freetown2> and you can set permissions no problemo?[04:06:50] <vraa> yes and no, i only had one user at the time, and nothing was tied into the win2k8 stuff, i believe permissions were all local based (ie: root and vraa)[04:07:01] *** Mace_N900 has quit IRC[04:07:17] <freetown2> maybe we should hit a private channel :P[04:07:20] <vraa> it's not a worry anymore, samba allows you to set permission per user, or per share, or etc[04:07:25] <vraa> no it's okay i have to run now[04:07:29] <vraa> thanks for the advice[04:07:33] <vraa> i will test sharemgr more, adios[04:07:42] <freetown2> adios[04:07:57] *** vraa has quit IRC[04:08:13] *** balrog has quit IRC[04:08:49] *** timeless_mbp has quit IRC[04:09:53] *** balrog has joined #opensolaris[04:14:29] *** Shoggoth has joined #opensolaris[04:15:07] *** sergiusens has joined #opensolaris[04:19:09] *** e1kg has quit IRC[04:19:09] *** balrog has quit IRC[04:20:04] *** eclark has left #opensolaris[04:20:55] *** balrog has joined #opensolaris[04:26:18] *** Mace_N900 has joined #opensolaris[04:28:17] *** balrog has quit IRC[04:29:57] *** balrog has joined #opensolaris[04:34:00] *** Mace_N900 has quit IRC[04:37:15] *** tg has quit IRC[04:38:16] *** tg has joined #opensolaris[04:39:00] <Shoggoth> hi[04:40:05] *** balrog has quit IRC[04:40:09] <freetown2> hello[04:40:58] *** balrog has joined #opensolaris[04:43:14] *** jthunder has left #opensolaris[04:46:23] <Shoggoth> I was just reading an entry on Darren Moffat's blog about lofi encryption (http://blogs.sun.com/darren/entry/opensolaris_disk_encryption_in_snv)[04:47:37] *** derekv has joined #opensolaris[04:47:39] <Shoggoth> I note that he mentions lofi performance not being very good.[04:48:35] <Shoggoth> Is anyone willing to comment on lofi's performance vs the linux equivalent - the linux loopback driver?[04:54:44] <CIA-21> Zhao Edgar Liu - Sun Microsystems <Edgar.Liu at Sun dot COM>: 6945724 audiohd path building code should be more clear[04:58:49] *** Havokmon has joined #opensolaris[04:59:14] *** sergiusens1 has joined #opensolaris[05:00:56] *** sergiusens has quit IRC[05:00:58] <Havokmon> I need the si3124 driver from 136 - What's the best way to get it? Or - when will 136 pkgs be available on pkg.opensolaris.org ?[05:10:25] *** Mousk4r1 has quit IRC[05:10:31] *** myosound has quit IRC[05:10:38] <Chipdancer> Havokmon: best way to get it, wait a month or so until it's all (expectedly) rolled out[05:11:01] <Chipdancer> Havokmon: other way, steal it from ON, compile it yourself[05:11:01] <Chipdancer> ([05:11:02] <Chipdancer> can't[05:11:03] <Chipdancer>[05:11:09] <Chipdancer> say how well that will work though)[05:15:48] <Havokmon> yeah I figured as much :( .. I just had to fall back to 118 - 134 just crashed twice in 2 hours due to that driver... hopefully 118 will behave as it did before... it was only crashing once every 90 days or so..[05:23:32] *** Mousk4r1 has joined #opensolaris[05:24:38] *** Havokmon has quit IRC[05:25:32] * freetown2 is glad he don't upgrade between dev releases.[05:25:50] <freetown2> Is the OpenSolaris support contract any good?[05:28:48] *** gm152 has quit IRC[05:32:18] *** randw has quit IRC[05:32:35] <rkeene> No.[05:38:46] <freetown2> never had the option for OpenSolaris support in HK anyway.[05:42:19] <rkeene> The people who bought it got the shaft.[05:43:03] *** alexc has joined #opensolaris[05:45:27] <freetown2> glad I don't have to be red-faced explaining things to management then[05:49:33] <Shoggoth> I've got 3 zpool's running on a paratirtual solaris VM, all three pools are encrypted, two of them are encrypted by the (linux) dom0 and the third is using lofi encryption. I've run a scrub on all three pools and the lofi encrypted one is taking _much_ longer than the others. Is lofi performance really that bad?[05:50:46] <Shoggoth> or should I go looking for another cause to the slow performance?[05:51:58] <Shoggoth> eh... typo in my question: s/paratirtual/paravirtual/ :)[05:54:50] <CIA-21> Evan Yan <Evan.Yan at Sun dot COM>: 6930155 pcitool fails with IO error on osol_133/x6440[05:55:03] *** Odin- has quit IRC[06:03:43] *** theRealballChalk has joined #opensolaris[06:07:10] <trochej> Coffee[06:12:08] <DesiJat> jamaica blue mountain[06:12:13] <DesiJat> bought a half-pound today[06:12:19] <DesiJat> will try it in the morning[06:12:27] <DesiJat> hope it's worth what i paid (doubtful)[06:14:39] <Shoggoth> anyone?[06:21:37] *** popster has joined #opensolaris[06:23:27] *** derchris has quit IRC[06:23:32] *** derchris has joined #opensolaris[06:35:59] <Jondice> interestingly, there is a coffee channel on freenode[06:39:23] *** theRealballChalk has left #opensolaris[06:39:30] *** popster has left #opensolaris[06:45:18] *** Mace_N900 has joined #opensolaris[06:45:27] *** IEEE has joined #opensolaris[06:45:44] *** IEEE has quit IRC[06:47:08] *** perlmongo has quit IRC[06:47:41] *** freakazoid0223 has joined #opensolaris[06:48:49] *** dec has quit IRC[06:49:16] *** dec has joined #opensolaris[06:50:09] *** p3rror has quit IRC[06:54:40] <CIA-21> Yu Wu - Sun Microsystems - Beijing China <Yu.Wu at Sun dot COM>: 6814017 mr_sas driver should support timeout[06:58:22] *** Zauberpo1y has quit IRC[06:59:53] *** Zauberpony has joined #opensolaris[07:00:08] *** vivid has quit IRC[07:02:02] *** Mace_N900 has quit IRC[07:07:31] *** vivid has joined #opensolaris[07:15:11] *** echobinary has left #opensolaris[07:15:18] <lewellyn> Jondice: is trochej an op? ;)[07:15:41] <Jondice> that was the first thing i checked, but no[07:15:56] *** lateau has joined #opensolaris[07:16:01] <winstonw> coffee[07:22:21] *** Gman has quit IRC[07:25:58] *** echobinary has joined #opensolaris[07:26:08] *** hudnix has quit IRC[07:28:47] *** sysspoof_ has joined #opensolaris[07:34:57] *** ivo_ has quit IRC[07:36:10] *** JonathanYC has quit IRC[07:37:39] *** sickness has quit IRC[07:37:39] *** alexc has quit IRC[07:41:55] *** hajma_ has quit IRC[07:42:15] *** solar_sea has joined #opensolaris[07:42:35] *** Rune^ has joined #opensolaris[07:43:36] <Shoggoth> how does one backup the contents of the pkcs#11 softoken ?[07:48:36] *** ThothCrimson has joined #opensolaris[07:55:29] <wereHamster> which CXXFLAGS do I need to set to enable POSIX.2, XPG4, SUS, SUSv2, SUSv3 ?[07:55:40] <wereHamster> (either of those standards)[07:56:14] *** hsp has joined #opensolaris[07:58:59] <madwizard> Coffee[07:59:00] <Triskelios> wereHamster: CPPFLAGS is the right place for these, and they're documented in standards(5) (the comments of sys/feature_tests.h might be quicker to read)[08:00:00] <wereHamster> man 3 getopt says it's defined in stdio.h or unistd.h, but I see the function is also defined in stdlib.h ..[08:04:08] <Triskelios> wereHamster: that's guarded by defined(__EXTENSIONS__) || (!defined(_STRICT_STDC) && !defined(__XOPEN_OR_POSIX)) so would be for compatibility (legacy code?) if standards conformance isn't needed[08:07:02] <wereHamster> well, I'm giving up..[08:10:40] <Triskelios> what are you having trouble with?[08:13:04] <wereHamster> http://trac.xapian.org/ticket/469[08:14:27] <lblume> /usr/lib/gcc ?![08:17:10] *** merzo has joined #opensolaris[08:25:08] *** Triskelios has quit IRC[08:25:56] *** brougham has quit IRC[08:26:14] *** eycel has joined #opensolaris[08:27:44] *** jnss is now known as ANNA[08:32:12] *** brougham has joined #opensolaris[08:35:08] *** Triskelios has joined #opensolaris[08:35:23] *** ozux has joined #opensolaris[08:36:19] *** phimic has joined #opensolaris[08:39:06] *** vivid has quit IRC[08:41:17] *** brougham has quit IRC[08:41:39] *** ingenthr has joined #opensolaris[08:46:02] *** brougham has joined #opensolaris[08:48:53] *** |AbsyntH| has joined #opensolaris[08:49:04] <Triskelios> wereHamster: try -ansi or -std=c++{98,0x}. your problem only occurs with g++ as _STRICT_STDC would normally be defined, but g++'s default enables extensions[08:50:11] *** relling has quit IRC[08:51:01] *** Yu\2 has joined #opensolaris[08:51:09] <Triskelios> wereHamster: or just follow http://trac.xapian.org/wiki/BuildInstallXapianOnSolaris or something. I notice SFExapian* exists, actually[08:51:23] *** vivid has joined #opensolaris[08:51:53] <Triskelios> wereHamster: yeah, just use SFExapian*, they're updated as of March[08:53:53] *** vivid has quit IRC[08:54:33] *** vivid has joined #opensolaris[08:54:46] <Triskelios> wereHamster: oh, your bug is against a different branch of Xapian than the ones in the spec (1.0.x)... stick with -std=/-ansi then or try to update the Studio specs[08:56:59] *** heldchen has quit IRC[08:57:20] *** heldchen has joined #opensolaris[08:58:52] *** Beelsebob has quit IRC[08:59:03] *** ozux has quit IRC[08:59:07] *** Beelsebob has joined #opensolaris[09:03:42] <madwizard> Guy, I wan to make a scp to a non-global zone. It's been very slow for some time and today I got[09:03:45] <madwizard> :[09:03:48] <madwizard> s/Gy/Gys/[09:03:49] <madwizard> Grr[09:03:52] <madwizard> Guys[09:04:10] <madwizard> No more processes.[09:04:26] <madwizard> Any thoughts on what could be a reason?[09:04:40] <madwizard> I'm thinkign that global zone is fsckd up, but I don[t have access there[09:04:55] *** kwinz3 has quit IRC[09:05:12] *** gebi has joined #opensolaris[09:05:53] *** Beket has joined #opensolaris[09:07:35] *** Erwann has joined #opensolaris[09:07:48] *** perlmongo has joined #opensolaris[09:17:12] <tsoome> you are out of memory.[09:19:02] <madwizard> ah[09:19:14] <madwizard> Possible[09:19:25] <madwizard> I have information, that global zone is pretty slow, too[09:19:51] <trygvis> if so iostat should show lots of swapping activity[09:21:38] <madwizard> Would I see it from nonglobal zone?[09:22:22] <trygvis> yep[09:22:41] <trygvis> hm, I think so at least. can't say I've tried[09:22:59] <CosmicDJ> IIRC you can display the max. num of processes with sysdef[09:24:40] <madwizard> yes, I can[09:25:40] *** Beelsebob has quit IRC[09:26:36] *** timeless_mbp has joined #opensolaris[09:26:36] *** timeless_mbp has joined #opensolaris[09:27:07] *** biant92 has joined #opensolaris[09:30:13] *** jamesd2 has quit IRC[09:30:31] *** jamesd2 has joined #opensolaris[09:32:44] *** twisti has joined #opensolaris[09:34:14] *** dalibor has joined #opensolaris[09:34:19] <FastJack> good morning[09:34:29] <asyd> \coin[09:34:34] <asyd> oups[09:34:36] <asyd> \_o<[09:35:32] *** Mousk4r1 has quit IRC[09:37:20] <Tonnerre> asyd, bizarre[09:37:40] <asyd> yeah, made a mistake with /coin (my alias for the duck, hehe)[09:37:58] *** Aria has quit IRC[09:38:11] <asyd> hmm I have a strange thing with 802.1q, while snooping a vlan device, I can see outgoing trafic, while I can't see incoming trafic[09:38:33] *** CodeWar has joined #opensolaris[09:38:45] *** CodeWar has quit IRC[09:39:13] *** CodeWar has joined #opensolaris[09:39:44] *** CodeWar has joined #opensolaris[09:42:28] *** Shoggoth has quit IRC[09:43:11] *** Shoggoth has joined #opensolaris[09:46:25] *** timsf has joined #opensolaris[09:46:48] <madwizard> Coffee[09:48:03] *** dalibor has quit IRC[09:49:46] <Shoggoth> madwizard: sun aught to have used you as the java mascot instead of duke... you must just about mainline caffeine :)[09:50:14] <madwizard> :)[09:50:34] *** jamesd2 has quit IRC[09:51:02] *** jamesd2 has joined #opensolaris[09:51:52] *** tsoome has quit IRC[09:52:02] <Shoggoth> has anyone experimented much with lofi encryption?[09:53:06] <CosmicDJ> still waiting for zfs crypto...[09:53:50] <Shoggoth> CosmicDJ: I hope you're not holding your breath... I better call an ambulance[09:54:50] <CIA-21> xun ni - Sun Microsystems - Beijing China <Xun.Ni at Sun dot COM>: 6818497 si3124 has no quiesce[09:54:51] <CIA-21> changqing li - Sun Microsystems - Beijing China <Changqing.Li at Sun dot COM>: 6838180 tcp corruption seen with northstar cards., 6925276 e1000g not consistent with link_lock usage, 6933844 NULL ptr deref in e1000g_rx_setup() due to inconsistency between recv_list & rx_desc_num[09:54:53] <CIA-21> George Wilson <George.Wilson at Sun dot COM>: 6911420 ZFS device removal detection should work with SCSAv3[09:55:39] <Shoggoth> I've got to ask: is CIA-21 a bot? What gives with his/it's posts?[09:55:51] *** Beelsebob has joined #opensolaris[09:56:07] <CosmicDJ> http://blogs.sun.com/darren/entry/encrypting_zfs_pools_using_lofi[09:56:53] <taemun> is there a way to get a copy of a file that was just overwritten on a zfs pool?[09:56:54] *** hajma has joined #opensolaris[09:57:00] <taemun> (no snapshots, no anything intelligent)[09:57:04] <timsf> Shoggoth: it watches commits to the ON source tree[09:57:08] <Shoggoth> CosmicDJ: yep read that...problem is that performance is _awful_[09:57:18] <trygvis> Shoggoth: http://cia.vc/stats/project/opensolaris[09:57:32] <Shoggoth> timsf: right, I guessed it might be something like that... but wasn't 100%[09:58:04] <timsf> taemun: not without digging around with zfs internals - and even then, it's hard.[09:58:12] <taemun> :([09:58:16] <taemun> oki thanks timsf[09:58:21] *** CodeWar has quit IRC[09:59:33] <Shoggoth> CosmicDJ: I've currently got 1 pool setup that way and two others encrypted using the linux crypto stuff (my solaris box is a paravirtual VM)[10:00:04] <Shoggoth> and lofi is about an order of magnitude slower... which I'm finding very surprising[10:00:25] <Shoggoth> I'm wondering if there's something that I've setup incorrectly or is it the case that lofi is just really crap[10:00:29] <CosmicDJ> you're running opensolaris on virtual hardware?[10:00:33] <Shoggoth> yep[10:00:42] <CosmicDJ> did you try bare metal?[10:00:43] <timsf> http://markmail.org/message/jqest4f2yrbvbhxj?q=recovery+deleted+files+zfs+list:org%2Eopensolaris%2Ezfs-discuss&page=1 has more about recovering deleted files[10:01:22] <Shoggoth> CosmicDJ: I'm thinking of trying it out with a LiveCD... but currently two of my pools are encrypted via linux...[10:01:23] <timsf> (och, there's more replies than that - hm )[10:02:15] <Shoggoth> CosmicDJ: the odd thing is that running encryption in the dom0 has almost zero impact...[10:02:18] <taemun> cheers timsf[10:02:26] <taemun> (you're the third "tim<tab>" lol[10:02:27] <taemun> )[10:02:33] <Shoggoth> that said neither the dom0 nor the domU is CPU or memory bound[10:03:13] <timsf> life's hard.[10:04:18] <madwizard> Coffee[10:04:23] <Shoggoth> again? lol[10:04:26] *** lucalenardi has joined #opensolaris[10:04:42] <Shoggoth> maybe madwizard is also a bot ;)[10:04:51] <oninoshiko> let the poor man have his vice in peace[10:05:18] <Shoggoth> hey I got no problem with it... I'm in awe of it[10:05:19] <Shoggoth> :)[10:10:53] * causality needs a new thinkpad[10:11:05] <causality> any of you guys got an x60/x200/x300?[10:12:47] <taemun> an x60 would be a new one for you?[10:12:47] <taemun> lol[10:13:25] * Shoggoth knows nothing of thinkpads[10:14:17] *** timsf1 has joined #opensolaris[10:14:25] <causality> taemun: i have a 240 and x30 :)[10:14:36] <taemun> hah[10:14:47] <causality> the 240 is the smallest thinkpad ever made[10:14:49] <taemun> what did you want to know?[10:15:03] <causality> things like build quality[10:15:14] <causality> i believe the x60 was the last proper thinkpad, according to wikipedo[10:15:20] *** timsf has quit IRC[10:15:57] *** man_with_a_stic1 has joined #opensolaris[10:16:03] *** devians has joined #opensolaris[10:16:11] <taemun> friend has an x300[10:16:13] <taemun> its nice[10:16:22] <Shoggoth> eh... not much... I know they're renowned for their build quality but I can obtain HP hardware at staff price so a thinkpad is off my menu[10:16:25] <taemun> I can't really compare it with thinkpads of old though[10:16:32] <man_with_a_stic1> does anyone know how to run a cronjob every x minutes with osol?[10:16:40] <taemun> my hp 2730p (tablet) is nice enough too[10:16:50] <taemun> although the finish isn't exactly what you'd call ultra durable[10:16:58] <Shoggoth> I'm looking at getting one of the new 8440w elitebooks[10:17:01] <timsf1> man_with_a_stic1: you can't (easily)[10:17:13] <Shoggoth> but they're not available outside of north america yet :-|[10:17:15] <taemun> unless x = 60[10:17:19] *** timsf1 is now known as timsf[10:17:30] *** dalibor has joined #opensolaris[10:17:39] <man_with_a_stic1> */2 doesnt work huh?[10:17:48] <timsf> old cron impl[10:18:07] *** ewdafa has joined #opensolaris[10:18:09] <causality> taemun: i think the x300 is probably out of my price range[10:18:14] <timsf> You can jump through hoops and create an at-job which reschedules itself though[10:18:18] <causality> the disk failed in my x30 so perhaps i should just replace that.[10:18:26] <causality> i wonder how well opensolaris would run on an x30[10:18:29] <Shoggoth> probably[10:18:30] <timsf> or do what I did in the auto-snapshot bits[10:18:34] <taemun> causality: the x300 is legacy by now[10:18:36] <timsf> but that sucked too.[10:19:09] <taemun> oh you're *that* timsf lol[10:19:17] <man_with_a_stic1> damn, thanks[10:19:59] <timsf> Yeah, sorry[10:20:28] <causality> taemun: hm![10:20:40] <taemun> is the zfs/backup field in the current implementation supposed to be functional?[10:21:04] <timsf> yeah, it should work[10:21:05] <causality> taemun: the x300s was the best X series[10:21:15] <causality> still selling for $1k+ used[10:21:16] <taemun> causality: the x200 is a chunky x200s (slim)[10:21:22] <taemun> x300s?[10:21:28] <taemun> its an x300 or an x301[10:21:29] <timsf> there's a psarc case at the moment for the new timesliderd implementation to do backup[10:21:35] <causality> taemun yeah i meant 302[10:21:38] <causality> 301 :([10:21:57] <timsf> PSARC/2010/129[10:22:10] <taemun> well I'm stuck on tablets, so it'd be an x201t or null :P[10:22:22] <taemun> timsf: reason I ask is because I literally couldn't get it to do anything[10:22:26] <taemun> it wouldn't produce errors[10:22:26] <Shoggoth> hey does anyone know how to backup the pkcs11 soft token?[10:22:32] <taemun> and it wouldn't do anything[10:22:36] <taemun> thats probably my fault, but anyway[10:22:50] <taemun> ended up going with a custom script in cron[10:22:57] <timsf> Huh, worked for me. Probably worth debugging on the command-line, or wait for http://arc.opensolaris.org/caselog/PSARC/2010/129/time-slider-phase2-one-pager.txt[10:23:05] *** lucalenardi has quit IRC[10:23:15] <taemun> was just finding that, thanks[10:24:02] <timsf> imho this new impl. it still doesn't cut it, but I've been over that with the project team - not my problem anymore.[10:24:14] <taemun> :([10:24:14] *** easyE has quit IRC[10:29:25] *** jaywalk has quit IRC[10:29:26] *** trygvis has quit IRC[10:29:28] *** trygvis has joined #opensolaris[10:30:05] *** commander has quit IRC[10:31:28] *** commander has joined #opensolaris[10:32:03] *** tsoome has joined #opensolaris[10:33:09] *** jaywalk has joined #opensolaris[10:36:57] *** houst0n has joined #opensolaris[10:37:15] *** commander has quit IRC[10:37:19] *** commander has joined #opensolaris[10:38:08] *** hsp has quit IRC[10:38:45] *** hsp has joined #opensolaris[10:39:43] *** hsp has joined #opensolaris[10:40:10] *** alexc has joined #opensolaris[10:40:23] *** pegasus-rider has joined #opensolaris[10:40:50] *** hsp has quit IRC[10:41:28] *** hsp has joined #opensolaris[10:41:32] <pegasus-rider> I'm wondering if OS 2009.06 Live CD supports Firewire networking? The was no such option in GUI, may ifconfig be helpful?[10:42:12] <pegasus-rider> Also prompt please if there's HFS+ support?[10:42:20] *** hsp has quit IRC[10:42:24] <asyd> s 18[10:42:26] <asyd> oups[10:42:46] <houst0n> You can get HFS+ support via fuse[10:42:57] <pegasus-rider> ... And FUSE via IPS...[10:43:04] *** hsp has joined #opensolaris[10:43:10] * houst0n shrugs[10:44:38] *** AukeF has joined #opensolaris[10:45:59] <lblume> What is the point of FireWire networking?[10:46:53] <pegasus-rider> Well, first of I can't get my (winning :-) ) Ethernet NIC working under OS[10:47:21] *** RoyK has joined #opensolaris[10:47:34] *** hsp has quit IRC[10:47:51] <pegasus-rider> Then there's speed. Gigabit Ethernet rules, just you have to get switches et al that support it.[10:47:53] *** ma3x has quit IRC[10:48:25] <houst0n> Go go infiniband![10:48:42] <pegasus-rider> Yeah, Light peak takes over yon ould world :-)[10:50:04] <RoyK> 10Gbps is nice too[10:51:24] <lblume> Gb switches are so cheap nowadays.[10:51:32] *** gebi has quit IRC[10:51:50] <lblume> And for back to back like FW, you don't even need one.[10:52:22] <pegasus-rider> Well, I've got an Apple machine with an added bonus of FW NAS mode[10:54:05] <lblume> If you've got money to waste in a Mac, you won't even notice the price of a Gb switch on the bill.[10:54:30] <pegasus-rider> I've got Mac for free, so I didn't noticed it :-)[10:55:35] <lblume> So you saved money, and can afford more stuff, still no excuse! ;-)[10:55:37] <tsoome> lblume: i have wasted more money on pc than on mac:P[10:56:09] <oninoshiko> tsoome: me too, but I dont own any macs.[10:56:30] <pegasus-rider> oninoshiko: you lucky you don't ;-)[10:56:38] <tsoome> that was before i learnt that if you wanna buy cheap and good, you need to buy twice.[10:57:15] *** hsp has joined #opensolaris[10:57:17] <lblume> Cheap and good means time wasted to look carefully at what you're buying.[10:57:56] *** man_with_a_stic1 has left #opensolaris[11:02:14] *** KermitTheFragger has joined #opensolaris[11:05:25] *** milek has joined #opensolaris[11:05:58] *** Fallen_Demon has joined #opensolaris[11:08:33] *** ewdafa has quit IRC[11:12:29] *** panos has quit IRC[11:18:38] *** ewdafa has joined #opensolaris[11:22:03] *** betchou has joined #opensolaris[11:22:10] *** jamesd2 has quit IRC[11:22:37] *** jamesd2 has joined #opensolaris[11:23:25] <madwizard> Coffee[11:24:33] <Shoggoth> CosmicDJ: are you still about? I tried running on bare metal and it's still as slow as hell[11:27:00] *** purserj has quit IRC[11:27:17] *** purserj has joined #opensolaris[11:27:37] *** stewart has quit IRC[11:27:53] *** gebi has joined #opensolaris[11:29:25] *** dalibor has quit IRC[11:32:52] *** Shoggoth has quit IRC[11:40:03] *** Shoggoth has joined #opensolaris[11:40:37] <Shoggoth> what would be the impact of killing the power during a zfs scrub?[11:43:09] <Shoggoth> Is there any risk?[11:44:26] *** dalibor has joined #opensolaris[11:45:00] <tsoome> scrub is read only activity. but killing power while system is active is *always* dangerous.[11:45:43] <jbit> you could always just stop the scrub and shutdown :)[11:46:05] <Shoggoth> eh... I can't stop it[11:46:11] <dnaumov> sure you can[11:46:15] <dnaumov> man zpool[11:46:15] <Shoggoth> damn thing keeps core dumping[11:46:50] <Shoggoth> yes... yes... I know.... zpool scrub -s <pool>.... the machine is toast... I can't even do an ls without a core dump[11:47:10] *** tosh has joined #opensolaris[11:47:32] <tsoome> how much ram you have and what build?[11:47:34] <Shoggoth> tsoome: that's what I wanted to know... I'm guessing it does write occasionally but only when correcting an error[11:47:39] <Shoggoth> it's a livecd[11:48:24] <Shoggoth> I normally run as a paravirt VM but I was having a performance issue so I ran a test "bare metal" to make sure it wasn't the hypervisor that was giving me grief[11:48:38] <tsoome> what build?[11:48:53] <Shoggoth> b129[11:49:05] <tsoome> why so old?[11:49:30] <Shoggoth> someone here recommended it... I tried out b134 but I found it to be a little on the unstable side[11:49:41] <Shoggoth> it works fine when I'm running it as a VM[11:49:42] <tsoome> i see[11:49:54] <Shoggoth> it's just this instance with the livecd... murphy's law I guess[11:50:06] *** fleebailey33 has joined #opensolaris[11:50:07] *** noyb has quit IRC[11:50:39] <tsoome> can be, also the livecd is not exactly perfect for any kind of perfomance tests...[11:50:42] <Shoggoth> I'd let the scrub run to completion except that's what I was trying to test[11:50:42] <AukeF> speaking of old releases, I have not seen any options to migrate from pre-IPS (<=nv130) to post-IPS releases, which could also be a reason why people are stuck with old versions[11:51:04] <tsoome> AukeF: backup and restore.[11:51:07] <Shoggoth> I'm using lofi encryption on the zpool and the scrub is insanely slow[11:52:38] <AukeF> tsoome, ie. manually putting back changes in /etc/default/ and /etc/system, as well as other configuration files, *shudders*[11:52:48] *** twisti has quit IRC[11:53:02] <tsoome> you really should not have anything in /etc/system[11:53:35] <AukeF> i fully agree. unfortunately I have an LSI SAS mpt driver[11:53:39] <tsoome> if you do, you either have no idea what you are doing or you have no problem with recovering those changes.[11:53:54] <Shoggoth> tsoome: so I know that there's always a risk with shutting down a live box, but you're sure that scrubbing is mostly reads?[11:54:29] *** twisti has joined #opensolaris[11:54:31] <joshua_> ZFS should be resilient against hard power-offs by its design[11:54:41] <CIA-21> sonam gupta - Sun Microsystems - Bangalore India <Sonam.Gupta at Sun dot COM>: 6922197 Remote printing broken if banner page is disabled[11:54:42] <CIA-21> sonam gupta - Sun Microsystems - Bangalore India <Sonam.Gupta at Sun dot COM>: 6941553 Banner Title field switch to standard input if the source file path is longer as 23 characters[11:54:43] <tsoome> scrub is reading data and verifying checksum. if it detects mismatch, and you have redundancy, zfs will repair the data block[11:54:44] <CIA-21> gowtham thommandra - Sun Microsystems - Bangalore India <Gowtham.Thommandra at Sun dot COM>: 6882563 svc:/system/filesystem/local should re-run swapadd[11:55:29] <Shoggoth> ok... that's what I thought but I appreciate the reassurance... cheers![11:55:41] <joshua_> tsoome, my impression is that *all* things go through the transaction layer, and so it can't directly repair the data block -- just set up a repair to be done at the next txg commit?[11:55:46] <tsoome> whole point of scrubbing is to make sure all data blocks are checked, as in normal disk usage only little % of data is in actual use...[11:56:36] <Shoggoth> I know one thing is for sure... I'll be dumping lofi[11:56:55] <tsoome> how much ram you have there?[11:57:00] <Shoggoth> 8GB[11:57:41] <Shoggoth> quad core 2.4GHz[11:58:10] <tsoome> not bad. ofc if encryption is in use, it can depend a lot which build is it etc[11:58:20] <CosmicDJ> did you compare openssl's speed test on both OSs?[11:58:31] <Shoggoth> ofc?[11:58:34] <tsoome> if its using crypto framework or its own implementation etc[11:59:07] <Shoggoth> ahh... right[11:59:18] <Shoggoth> CosmicDJ: no I haven't tried that... I might do that after I reboot[11:59:55] <Shoggoth> would b134 be using the crypto framework?[12:00:25] <tsoome> no idea, need to check change logs[12:00:47] <Shoggoth> isn't that the latest build?[12:00:47] <dnaumov> ssh 192.168.1.126[12:00:50] <dnaumov> oops[12:00:50] <dnaumov> :D[12:00:55] <Shoggoth> :)[12:01:11] <tsoome> im not really sure if they do work on lofi or its just an "proof of concept" implementation[12:01:12] <tsoome> w[12:01:44] <tsoome> as most people expect zfs native crypto anyhow:)[12:02:14] <Shoggoth> tsoome: I'm wondering the same... I was planning on migrating across and running solaris bare metal, currently my other two zpool's are encrypted in the dom0 via linux device mapper and they're an order of magnitude faster[12:03:04] <tsoome> can be, solaris is server os and filesystem encryption in servers is something you dont exactly need to have[12:03:33] <tsoome> for servers there are other methods to ensure security....[12:03:43] <Shoggoth> mmmm.... I don't know that I'd agree with that... I've worked on quite a few sites where it's used extensively[12:04:01] <Shoggoth> granted it's a minority[12:04:19] <Shoggoth> but a significant one[12:05:41] <tsoome> if you cant trust physical security of your server room, then you have way bigger problems.[12:06:09] <houst0n> I've never seen a production machine w/ encrypted disks[12:06:13] <tsoome> obviously there are still some well justified use cases[12:06:17] <Shoggoth> tsoome: I would agree somewhat.. but some sites are rather complicated[12:06:20] <houst0n> Laptops, sure.. Server? No, not a good idea[12:06:22] <tsoome> agree[12:06:39] <Shoggoth> and besides some sites don't do it because it's a good idea but because of policy from on high[12:06:43] <tsoome> well say your "server" is in truck and mobile.[12:06:58] <tsoome> physical security is issue then[12:07:01] <Shoggoth> well yes... some of the "sites" I've dealt with have those as well[12:07:02] <Shoggoth> :)[12:08:04] <Shoggoth> in any case I'm a little surprised... my normal experience with anything solaris is that it tends to perform better than the linux equivalent[12:08:17] <Shoggoth> oh well! I guess you can't win 'em all[12:08:30] <tsoome> fs level encryption is very new and it is experimental[12:09:07] <tsoome> you wont even see it in real solaris - its not even in solaris 10[12:09:44] <Shoggoth> yes... and I agree somewhat with the zfs team's decision to do encryption at the file level. however I still believe end users aught to be provided with a choice to use full-disc-encryption rather than file level encryption[12:10:11] <Shoggoth> FDE has been around for quite some time and is a well proven technology[12:10:47] <tsoome> its told to be coming, but not there yet. it does need some more technologies to be integrated in zfs to function properly[12:10:56] *** p3rror has joined #opensolaris[12:11:11] <Shoggoth> yes... file level encryption that is... but that's not the same as FDE[12:11:30] <tsoome> tbh... i dont believe in FDE. its ugly hack.[12:11:45] <Shoggoth> I disagree... it's all about risk management[12:11:57] <Shoggoth> _both_ are useful tools to have in your toolbox[12:12:43] <tsoome> i dont argue that crypto is useful, but those encryption implementations are technically quite ugly.[12:12:53] <Shoggoth> eg?[12:13:39] <nikolam> WOW. I turned off months ago automatic snapshots making. And i forgot to remove some OLD snapshots.[12:13:46] <Shoggoth> lol[12:13:58] <Shoggoth> now you can cancel your PO for more discs :)[12:14:00] <nikolam> This morning I went to that GUI and.. reciever my 20GB back[12:14:16] <nikolam> isnt that nice..[12:14:38] *** chrismat1 has joined #opensolaris[12:15:33] <Shoggoth> tsoome: thanks for your input... I'm going to reboot that machine now... I have to head off.[12:15:36] <Shoggoth> thanks![12:16:03] *** |AbsyntH| has quit IRC[12:18:23] *** calumb has joined #opensolaris[12:18:58] *** betchou has quit IRC[12:19:42] *** biant92 has left #opensolaris[12:20:30] *** betchou has joined #opensolaris[12:22:35] *** Shoggoth has quit IRC[12:25:05] *** sergiusens1 has quit IRC[12:25:08] <chrismat1> I cant # pkg refresh[12:25:22] *** calumb is now known as calAFK[12:25:35] <chrismat1> it says can't resolve http://pkg.opensolaris.org/dev/catalog/0[12:25:48] <chrismat1> is that problem local or general?[12:26:03] <chrismat1> running snv_132[12:26:04] <seanmcg> can you ping pkg.opensolaris.org ?[12:26:26] <chrismat1> I can ping pkg.opensolaris.org[12:26:41] <chrismat1> actually can't[12:26:43] <chrismat1> hmmm[12:27:14] *** mroconnor has joined #opensolaris[12:27:56] <chrismat1> can you ping pkg.opensolaris.org?[12:28:19] <chrismat1> I cant ping it from my DSL connection either[12:28:27] <jbit> Pinging pkg-na-1.opensolaris.org [72.5.123.32] with 32 bytes of data:[12:28:28] <jbit> Request timed out.[12:29:17] *** alexc has quit IRC[12:30:47] <chrismat1> have it ever answered on ping?[12:31:34] *** xtrondo has joined #opensolaris[12:32:49] *** timeless_mbp has quit IRC[12:32:54] *** dalibor has quit IRC[12:33:35] <Stric> ping isn't a way to verify if a web server works or not[12:34:13] <Stric> it's a way to see if there's something responding to ICMP Echo Request or not[12:34:27] <CosmicDJ> I'd still be nice if opensolaris.org, sun.com etc. would respond to a ping req.[12:34:30] <Stric> which some "security people" think is a "security risk"[12:35:39] <houst0n> Not really security, disabling it does pretty much 1/2 the impact of a ddos though[12:36:25] <RoyK> Stric: ping is a good way to see if a host is alive unless some paranoid sysadmin/security freak has blocked ICMP Echo Request/Reply. Some even blocks all of ICMP 'just in case' even though it's an important protocol meant to be used with IP[12:37:11] <Stric> RoyK: So it's not a good way to see if a host is alive then..[12:37:30] <Stric> because there are quite a few paranoid sysadmin/security freaks that blocks ICMP..[12:40:01] <Stric> Also, even if the host is not filtering ICMP and responds nicely, doesn[12:40:14] <Stric> doesn't mean that the web service that was actually requested works..[12:40:24] <RoyK> no...[12:40:51] <Stric> So in this case, heading to http://pkg.opensolaris.org with a web browser is a much better method[12:40:57] * RoyK uses nagios/icinga service checks for service control[12:41:09] *** betchou has quit IRC[12:42:53] <houst0n> I wish nagios would get a better webui[12:42:58] <houst0n> it's fucking ugly[12:43:05] <hali> icingas new web gui is nicer[12:43:11] <hali> (icinga is a nagios fork)[12:43:16] <houst0n> oh really?[12:43:19] <RoyK> yeah[12:43:21] <madwizard> http://icinga-web.icinga.org/icinga-web/[12:43:30] <RoyK> some more CSS makeup[12:43:35] <RoyK> but mostly the same[12:43:35] <Beket> and icinga should get another name[12:43:36] <Beket> :P[12:43:42] *** robinbowes has quit IRC[12:43:44] <RoyK> :)[12:44:08] <houst0n> I have 'redesign our monitoring platform' on my todo list for this week[12:44:18] <houst0n> You guys use icinga in prod? Any issues with it?[12:44:35] <hali> the new web ui is still beta[12:44:52] <hali> the standard icinga web ui is just the normal nagios web ui with a nicer css[12:45:06] <hali> so i'd be a little hesitant to use it in prod[12:45:09] <houst0n> Well, I have about 7-800 services defined[12:45:25] <houst0n> I would just run both until I was happy[12:46:08] *** robinbowes has joined #opensolaris[12:46:29] *** nickw has quit IRC[12:55:24] *** calAFK is now known as calumb[12:55:51] *** scottdickson has joined #opensolaris[13:03:04] *** lateau has quit IRC[13:03:06] <taemun> RoyK: I guess the only question I'd ask is are the SAS drives dual-interface[13:03:27] <taemun> (with reference to your email to zfs-discuss([13:04:07] <RoyK> afaics zfs list / df / du shows disk space used after compression - is there a good way to show the size of the data before compression when it's already on zfs?[13:04:22] <RoyK> taemun: what do you mean dual-interface?[13:04:36] <taemun> umm[13:04:49] <taemun> can talk to two sas interfaces at once, for fail-over[13:04:59] <taemun> dual port? I think is the correct term[13:05:20] <RoyK> ok[13:05:39] <RoyK> doesn't look like he needed that, though[13:05:47] <taemun> nah[13:05:58] <taemun> but I think that would be the only reason to buy 7200rpm SAS drives[13:06:07] <taemun> unless you just REALLY like having a "sas" sticker on them :P[13:06:15] <RoyK> :)[13:06:45] *** calumb has quit IRC[13:07:09] <RoyK> heh - this zfs send/receive will probably take ~14 days[13:07:19] <taemun> lol[13:07:28] <taemun> sounds like someone needs to invest in some 10GE[13:07:55] <RoyK> not really - the receiving machine uses 100% cpu for two zfs receive processes[13:07:59] <taemun> 0.o[13:08:01] <RoyK> old Opteron 1800[13:08:04] <taemun> ahhh[13:08:05] <taemun> right[13:08:10] <RoyK> ssh eats most of the cpu[13:08:16] <taemun> compression=gzip?[13:08:17] <taemun> oh[13:08:25] <seanmcg> RoyK, zfs list, man zfs to see the compression ratio used.[13:08:26] <RoyK> just normal compression[13:08:31] <seanmcg> I think[13:08:54] <RoyK> seanmcg: zfs show compressratio, yes, but it'd be nice to see that for, say, a directory tree[13:09:10] <RoyK> du --someflag[13:09:12] <RoyK> for instance[13:09:13] <CosmicDJ> RoyK: if that's in your internal network, why not just use nc instead of ssh?[13:09:25] <seanmcg> RoyK, log an RFE, patch it :)[13:09:38] <RoyK> seanmcg: heh[13:09:42] *** robinbowes has quit IRC[13:10:51] <RoyK> CosmicDJ: well, yeah, but it's been running for a day already and it will run for days anyway, so I guess I'll just wait[13:11:05] *** jamesd2 has quit IRC[13:11:36] *** jamesd2 has joined #opensolaris[13:14:05] <Gugge> RoyK: you could remove ssh from the send/receive path :)[13:14:09] <Gugge> and use something like netcat[13:14:12] <Gugge> its much faster :)[13:14:33] *** ziyourenxiang has joined #opensolaris[13:15:28] <madwizard> Coffee[13:16:50] *** swift has joined #opensolaris[13:19:09] *** sipior has joined #opensolaris[13:22:19] <RoyK> Gugge: I know, but it's running already and, as I said, it'll take several days anyway, so why bother :þ[13:34:05] *** hsp has quit IRC[13:34:36] *** hsp has joined #opensolaris[13:36:24] *** hsp has quit IRC[13:37:40] *** Mousk4r has joined #opensolaris[13:40:10] *** swift_ has joined #opensolaris[13:41:24] *** hsp has joined #opensolaris[13:41:55] *** _madmax_ has joined #opensolaris[13:42:11] *** Mick27 has joined #opensolaris[13:42:34] <FastJack> mhh, nice weather in poland. feels almost like a vacation (if I ignore the fact that I'm trapped in an office)[13:42:48] *** swift has quit IRC[13:43:03] *** swift_ is now known as swift[13:55:39] <madwizard> FastJack: Which part of Poland? :)[13:56:10] <FastJack> madwizard: szczecin[13:56:20] <FastJack> not quite. but close enough[13:58:38] *** seanmcg has quit IRC[13:59:44] *** seanmcg has joined #opensolaris[13:59:47] <madwizard> Hah![13:59:57] <madwizard> My city :)[14:01:16] <houst0n> Good weather in holland atm also[14:01:24] <houst0n> If I forget I'm trapped in an office, that is...[14:01:25] <houst0n> :P[14:02:30] *** niq has joined #opensolaris[14:02:30] *** niq has joined #opensolaris[14:03:20] *** sponix has quit IRC[14:07:45] *** ozux has joined #opensolaris[14:11:29] *** pablomh has joined #opensolaris[14:11:37] *** p3rror has quit IRC[14:15:46] *** _madmax_ has quit IRC[14:16:17] *** sk0rd has joined #opensolaris[14:17:38] *** forquare has quit IRC[14:18:54] *** forquare has joined #opensolaris[14:21:25] *** stewart has joined #opensolaris[14:25:22] *** p3rror has joined #opensolaris[14:26:17] <madwizard> :)[14:26:24] <madwizard> Coffee[14:32:32] *** bklang has joined #opensolaris[14:33:00] *** robinbowes has joined #opensolaris[14:43:45] *** Fallen_Demon has quit IRC[14:44:07] *** swift_ has joined #opensolaris[14:46:54] *** swift has quit IRC[14:47:01] *** swift_ is now known as swift[14:47:42] *** timeless_mbp has joined #opensolaris[14:47:43] *** timeless_mbp has joined #opensolaris[14:47:51] *** timeless_mbp has quit IRC[14:52:32] *** robo has joined #opensolaris[14:52:42] *** timeless_mbp has joined #opensolaris[14:52:43] *** timeless_mbp has joined #opensolaris[14:53:46] *** ozux has quit IRC[14:53:55] *** kmays has joined #opensolaris[14:54:07] *** sysspoof_ has quit IRC[14:56:17] *** Dagobert has joined #opensolaris[14:57:08] <kmays> If you're trying to build snv_139 (I'm using snv_137 now)... then read: http://hub.opensolaris.org/bin/view/downloads/on ... snv_137 has ZFS rev. 24[14:57:16] *** Andrew3 has quit IRC[14:58:47] *** twilling has joined #opensolaris[15:01:36] <jbit> hrm, revision 24 isn't listed on the zfs community page :([15:01:55] <jbit> ah actually it is, it's just not in the list[15:03:41] <kmays> If clustering...you can use Sun Cluster but you're on your own.[15:03:49] <kmays> ZFS rev 24: http://hub.opensolaris.org/bin/view/Community+Group+zfs/24[15:07:36] *** SunTzuKDE has joined #opensolaris[15:09:35] <kmays> If you use any of the optional OSOL distros like Belenix, you will want to have at least snv_128a for review. Nexenta has one of the latest and better distro CD releases at the moment to review the kernel putbacks..[15:10:56] <kmays> They labeled their patched kernel release as snv_134c which I think is running ZFS rev 22.[15:11:59] *** vivid has quit IRC[15:12:56] *** Spencer_tt has joined #opensolaris[15:13:50] *** Odin- has joined #opensolaris[15:15:25] *** thepickle has joined #opensolaris[15:16:01] *** pablomh has quit IRC[15:21:58] *** p3rror has quit IRC[15:22:37] *** synegy34 has joined #opensolaris[15:23:46] *** mroconnor has quit IRC[15:26:07] *** nikolam_ has joined #opensolaris[15:26:47] *** longcat has joined #opensolaris[15:27:27] <longcat> darn still not released. any inside word on why it's taken a month after release? im preparing for the worst.[15:29:13] *** nikolam has quit IRC[15:29:19] *** nikolam_ has quit IRC[15:29:37] <CosmicDJ> longcat: http://www.c0t0d0s0.org/archives/6510-Future-of-OpenSolaris.html "Next update 1st half CY2010"[15:30:09] *** nikolam_ has joined #opensolaris[15:30:16] *** biant92 has joined #opensolaris[15:31:32] <longcat> CosmicDJ: that's a very specific answer... insightful and helpful, thanks[15:31:43] <RoyK> CosmicDJ: that's about a month away, not so bad[15:32:22] <longcat> still not quite a release from oracle since it's from Dan Roberts (Director of Solaris Product Management)[15:32:23] <CosmicDJ> longcat: that's the most official statement you can get ATM[15:32:23] <causality> what does CY stand for?[15:32:35] <CosmicDJ> causality: wikipedia[15:32:35] <kmays> You know that OSOL 2010.05 seems about right...that was the first release date two years ago.[15:34:15] *** p3rror has joined #opensolaris[15:34:48] *** Animal-X has joined #opensolaris[15:36:10] *** vivid has joined #opensolaris[15:36:17] <longcat> imo, the timing is too close to other speculation-rich events, coupled with no technical information to explain the delay still justifies a rather suspicious view of events. but i guess time will tell.[15:36:25] *** leito has joined #opensolaris[15:36:33] *** lo3Viej has joined #opensolaris[15:36:54] *** achester has quit IRC[15:37:33] <longcat> trying to fit in renaming of packages right before a release was a rather dubious decision also[15:38:11] *** Andrew3 has joined #opensolaris[15:38:31] *** kwinz3 has joined #opensolaris[15:39:20] *** xapiens has joined #opensolaris[15:39:50] <causality> CosmicDJ: CY stands for wikipedia? interesting[15:40:16] <CosmicDJ> yeah right[15:40:22] <CosmicDJ> you got it[15:40:27] *** nachox has joined #opensolaris[15:40:37] *** Odin- has quit IRC[15:41:23] *** tg has quit IRC[15:41:55] *** tg has joined #opensolaris[15:44:15] *** pablomh has joined #opensolaris[15:45:00] *** ANNA is now known as jnss[15:46:16] *** kwinz3_ has joined #opensolaris[15:46:37] *** kwinz3 has quit IRC[15:46:48] <Alasdairrr> Is there seriously not going to be another IPS release until 2010.1H comes out (rumour being June)?[15:47:02] <Alasdairrr> Build 134 is such an abortion[15:48:56] *** betchou has joined #opensolaris[15:49:20] <tsoome> ask oracle[15:49:33] *** forquare has quit IRC[15:49:39] <tsoome> temple of delphi, i believe?[15:49:40] *** Beelsebob has left #opensolaris[15:49:53] <CosmicDJ> yeah everyone mails steve jobs about apple stuff, why don't you just ask uncle larry? ;)[15:50:12] *** forquare has joined #opensolaris[15:50:40] <Alasdairrr> I'm beginning to think that the "Open" part of OpenSolaris is there for irony[15:50:53] <tsoome> ask oracle[15:51:37] <kmays> Alasdairrr: not June, but May.[15:51:56] <Tonnerre> I guess the open stands for the open questions concerning the future[15:52:18] <tsoome> as time is passing, im starting to like oracle less and less. seems like thay are killing solaris just like hp did kill tru64 etc.[15:52:23] <Alasdairrr> What makes you so sure kmays? I've heard it may (mind the pun) slip to June[15:52:31] *** nickw has joined #opensolaris[15:53:41] *** Mousk4r1 has joined #opensolaris[15:54:20] <kmays> Alasdairrr: Well, not officially stated...but technically..that type of 35-day delay would be horrific.[15:54:35] *** lo3Viej has quit IRC[15:54:39] *** richb has joined #opensolaris[15:54:52] *** Burana has quit IRC[15:55:04] *** Mousk4r has quit IRC[15:56:55] <longcat> tsoome: how are they killing solaris?[15:57:30] <tsoome> have you seen *any* reasonable announcements from oracle regarding to solaris/opensolaris?[15:58:12] <tsoome> any announcement regarding to delays of next release?[16:00:25] *** pablomh has quit IRC[16:01:34] <longcat> that last point was the strongest signal of something wrong[16:02:15] <kmays> Oracle made statements alluding to Oracle Solaris 11 by Sep 2010...[16:02:40] <kmays> snv_139 is up on src.*[16:02:59] <kmays> (don;t know if that is final or just a marker.... I'm using snv_137 whic works OK for now.[16:03:16] <kmays> which[16:04:01] <kmays> Nexenta has snv_134c (their snv_134 + putback patches) which I'm using and also works very well for now.[16:04:05] <causality> managing an enterprise portfolio isn't as simple as simply announcing something on a mailing list[16:04:07] <causality> this ain't lunix.[16:04:13] <kmays> Right[16:04:28] *** psychicist has joined #opensolaris[16:05:16] *** thepickle has quit IRC[16:06:08] <kmays> So..you do have a distro that has a decent release of the kernel and snv_137 which does work... so I don;t think 35 days extra is really needed unless there is a MAJOR showstopper(s) which would cause such a delay or Oracle is just doing a CYA QA inspection for press release reasons..which is OK by me.[16:06:09] *** swift___ has joined #opensolaris[16:06:35] <kmays> Just off the cuff here.[16:07:18] <longcat> Yeah but without even a smoke screen of comfort 'we are having x issues, working them out' i perceive it as rather ominous[16:07:34] *** gebi has quit IRC[16:07:56] * jbit imagines it like a nasa shuttle launch, with lots of teams who need to "sign off" on the build[16:08:05] *** Mick27 has quit IRC[16:08:13] *** swift has quit IRC[16:08:16] <kmays> right..and legal crap.[16:08:34] *** perlmongo has quit IRC[16:10:05] <kmays> Nobody wants a black eye with a bad 'tested' release tossed out into the wild..just give them some time to make things right.[16:11:46] <tsoome> its 2 months overdue. and only thing we get is "oracle is not allowing to talk about it". its not about how much work needs to be done, its about attitude oracle is having against customers.[16:12:16] <kmays> People always get anxious whenever a new toy/baby/event is forthcoming but then gets delayed....it happens all the time..so this is nothing new.[16:12:20] <CosmicDJ> customers? so you pay for it?[16:12:22] <causality> it's a free product with no guarantees.[16:12:40] <causality> if you want accountability, pay for it.[16:13:50] <tsoome> CosmicDJ: not directly for opensolaris, no. but we do have quite some installations around and this company has been sun partner for about 15 years.[16:14:34] *** p3rror has quit IRC[16:14:41] <kmays> You also still have access to the latest public releases and sources...so you can really 'roll your own Live CD/DVD distro' if you really wanted to...or talk to Belenix/Nexenta/Milax..etc.[16:15:28] *** pgr has joined #opensolaris[16:15:29] <lblume> OpenSolaris /stable has no contract customers demanding to know when they get their update?[16:17:09] *** nikolam_ has quit IRC[16:17:12] <|woody|>[16:17:12] <|woody|> 05 Mar 2010[16:17:15] <|woody|> args[16:17:23] <tsoome> well, its little help from "demanding to know". its ready when its ready.[16:17:24] <|woody|> last stable update was on that date :)[16:18:01] *** nikolam_ has joined #opensolaris[16:19:40] <Alasdairrr> I think it's quite encouraging that people are so desperate to get their hands on 2010.03 - it clearly shows there's been a lot of great stuff go into OpenSolaris since 2009.06[16:19:58] *** Andrew3 has quit IRC[16:20:07] <lblume> tsoome: I was pointing the point to those people implying there are no customers for it :-)[16:20:24] *** Andrew3 has joined #opensolaris[16:21:15] <lblume> Alasdairrr: I'm certainly not desperate at all to get my hands on it. It's useless as you can't get any update for it if you don't have Oracle hardware.[16:22:16] *** tsoome has quit IRC[16:23:49] <kmays> lblume: I'd expect an Acct rep. would be around a supported customer to give them that Oracle approved "update" answer.[16:24:01] *** jthunder_ has joined #opensolaris[16:24:49] <kmays> lblume: but you are saying you can't get an update without Oracle hardware? Hmm.[16:25:38] *** jgracin has joined #opensolaris[16:26:06] *** leito has quit IRC[16:26:09] <lblume> Yes, you can't. Or some HP machines maybe, though my WAG is they are Solaris only. It's been posted all over the Oracle website and said by Oracle salespeople.[16:26:38] *** vraa has joined #opensolaris[16:27:11] *** astra-x has joined #opensolaris[16:27:28] <tehhobbit> http://tech.groups.yahoo.com/group/solarisx86/message/50926[16:27:48] <vraa> lol so i got the samba working, but my permissions are all messed up i think, every file i open ends up being in read only mode[16:28:18] *** biant92 has quit IRC[16:28:24] <lblume> vraa: Are there ACL's? Samba doesn't deal with those properly on ZFS.[16:28:45] <kmays> I know their hardware has priority (it runs here first)...just like Sun did... can't really be good to support HP/other hardware and it doesn't work on your own![16:28:53] *** swift has joined #opensolaris[16:29:32] <lblume> kmays: What are you talking about? I'm talking about *OpenSolaris* support (ie, *updates*), nothing to do with hardware. I'm not asking Oracle to fix my Asus mobo.[16:29:37] *** Shoggoth has joined #opensolaris[16:29:39] *** leito has joined #opensolaris[16:29:45] <kmays> hehe..ok.[16:29:51] *** Alasdairrr has quit IRC[16:30:13] <kmays> Yes..I'm talking about OSOL as well.[16:30:16] <lblume> But said updates are now tied to hardware, which is a problem for me.[16:30:25] *** gebi has joined #opensolaris[16:30:30] <longcat> opensolaris and solaris updates?[16:30:41] *** corretico has joined #opensolaris[16:30:45] *** swift___ has quit IRC[16:31:30] <lblume> Yes, they only deliver updates through entitlements, and only entitlements to Oracle hw.[16:31:39] <vraa> lblume: i'm not sure, how can i disable any permissions: i'd like everyone to be able to access read/write the files[16:31:39] *** p3rror has joined #opensolaris[16:31:58] <vraa> lblume: i copied all the files from a windows server 2k8 box[16:32:47] <lblume> vraa: permissions are wrong seen from where? On the osol server or from the windows clients?[16:32:54] *** biant92 has joined #opensolaris[16:33:49] <vraa> i think the windows clients, i would like to double check though, how can i verify on the osol box, i have putty open and logged in[16:33:50] *** Alasdairrr has joined #opensolaris[16:33:55] * alanc I should keep ignoring this channel so I don't have to correct all the absolutely wrong things I see in scrollback[16:34:07] *** swift_ has joined #opensolaris[16:34:08] <vraa> from what it appears: all files and dire belong to "vraa:staff"[16:34:17] <lblume> vraa: What does ls -l say?[16:34:33] *** pgr has quit IRC[16:34:38] <longcat> alanc: then make a crappy web page with the common misconceptions and your answers to it.[16:34:39] <jbit> alanc: you could just +m (moderate) the channel too and only allow oracle blessed people to talk :)[16:34:45] <longcat> so you dont have tp repeat yourself[16:35:00] <vraa> should i do the "ls -l" on my filesystem or in a specific folder?[16:35:04] <lblume> alanc: If anything I wrote is, please fell free to shame me publically, I'd rather be red-faced and corrected ;-)[16:35:25] <alanc> yes, but it's the uncommon stuff I see today - like "Sep 2010" - that's absolute bullshit, but I've never seen before[16:35:55] *** swift has quit IRC[16:36:10] <kmays> alanc: This was at a public presentation meeting...and it was not a SPECIFIC date..just timeline.[16:36:16] <vraa> lblume http://osol.pastebin.com/xP94Teis[16:36:22] *** randw has joined #opensolaris[16:36:23] <lblume> Sep 2010? For what? the stable release that's being clamored for by some?[16:36:43] <alanc> a) it's not even close b) no Oracle representative would give such a roadmap in public[16:37:07] <jbit> lblume: solaris 11 i guess[16:37:12] <lblume> vraa: Ouch, yes, the + indicates you've got ACL's all over the place, not good with Samba.[16:37:42] <vraa> lblume: i have been reading wikipedia on what ACL's are, how can i get rid of them all and start fresh with opensolaris?[16:37:56] <kmays> Oracle mention something about Solaris 11 during a presentaion and mention it was coming out very soon like Summer 2010... now if internally you have the NDA slapology going on...then thatis something else...but this was publically said.[16:37:59] <longcat> so uh if the updates are being tied to hardware, that could explain the delay as pkg is reworked to come up with a unique id for a system, and then building a framework for obtaining keys to be allowed access to the repository[16:38:39] <kmays> AlanC: Well..you are wrong there...they did...guess they should be shot then?[16:38:43] <lblume> longcat: I seriously doubt they'll do that. They don't do it for their DB, right?[16:38:59] <alanc> kmays: for lying to you? absolutely[16:38:59] *** kwinz3_ has quit IRC[16:39:01] <DerSaidin> vraa: man chmod and make sure your using the solaris one, not the gnu one[16:39:04] <kmays> hehe[16:39:16] <lblume> vraa: Start by having a look at chmod.[16:39:19] <longcat> < lblume> Yes, they only deliver updates through entitlements, and only entitlements to Oracle hw <-- I guess that answer was to solaris only then[16:39:24] <alanc> if they really said that and you didn't misunderstand it[16:39:30] <vraa> ah i think i did chmod -R 755 already[16:39:31] <kmays> alanc: hey, we know that is the real world for ya.[16:39:35] *** Burana has joined #opensolaris[16:39:36] *** merzo has quit IRC[16:40:06] <alanc> longcat: I don't know of anything like that being done for pkg, since pkg already has SSL certificates used to restrict access to the support contract repo[16:40:08] <longcat> or, that could mean updates as in the paid updates that was previously in place, versus having to use /dev[16:40:27] <lblume> longcat: What I mean is that you can only legally get updates for OpenSolaris and Solaris via an entitlement now. but it's still only a bunch of electrons that forms an email, and not a technical lock in.[16:40:36] *** kwinz3 has joined #opensolaris[16:41:12] <lblume> longcat: Yes, they are the same paid updates, only before, you could get access to them for any piece of hardware on the HCL. Not anymore.[16:41:37] <lblume> vraa: chmod -R 755 is *very* wrong[16:41:51] <vraa> lblume ya i think i am figuring out now, i am scouring the opensolaris.org discussion forums[16:42:02] <lblume> It turns everything into an executable,[16:42:11] <Stric> and doesn't remove any acls[16:42:48] <Alasdairrr> find /blah -type f -exec chmod 644 {} ';' ; find /blah type d -exec chmod 755 {} ';'[16:42:55] <Alasdairrr> (slightly less retarded)[16:43:27] <alanc> lblume: indirectly - the paid updates (i.e. S10 patches & OpenSolaris /support) require a support contract - getting a support contract requires an entitlement, which either comes with the hardware, or which you got by registering as you were supposed to when you downloaded S10 before the change-in-control[16:43:42] *** merzo has joined #opensolaris[16:43:59] <longcat> so when i downloaded and said "im using this for one computer" does that lock me in for my account?[16:44:17] <Stric> Alasdairrr: chmod -R u=rwX,og=rX /blah <- slightly shorter.[16:44:22] <lblume> alanc: This entitlement gets you access to perpetual use, but not to updates.[16:44:27] *** niq has quit IRC[16:44:31] <rkeene> alanc, Not that it matters anymore -- /supported is gone.[16:45:05] <alanc> I don't know if there is any way to tie the entitlement document to a specific computer or not[16:45:14] <Alasdairrr> It seems you can still buy dell and hp machines with Solaris 10 entitlements[16:45:38] <vraa> ah i see, i needed to do "chmod -R 777 /Home",m Alasdairrr I am trying to comprehend what you typed, what is your difference?[16:45:44] <Alasdairrr> but given Oracle's direction with Solaris you'd be insane to buy em until they clarify their plans[16:45:53] *** kwinz3 has quit IRC[16:46:03] <rkeene> alanc, It was done using PKI, so theoretically the smartcards the systems have could do it -- not sure if they have a PKCS#11 applet on them, though.[16:46:04] <Stric> vraa: If you think you needed to do that, then you don't understand the permissions.[16:46:05] <alanc> rkeene: /support should still be there, and they're still making us deliver updates to it - if it's offline, I'd expect it to be another temporary situation from the datacenter moves they've been doing to all OpenSolaris machines[16:46:20] <rkeene> alanc, It's been giving a 404 for like a month.[16:46:20] <lblume> alanc: Yes, it was always tied, to get this one you were supposed to give the SN of the computer you were dowloading Solaris for. So legally, for a single computer. But again, technically, nothing was binding.[16:46:42] <alanc> rkeene: have you reported it? that's the first I've heard anyone complain[16:47:03] <longcat> The requested URL /support was not found on this server. is pkg.opensolaris.org/support right?[16:47:12] <betchou> vraa: Stric: :D[16:47:17] <alanc> well, I did see someone complain they couldn't reach pkg.sun.com because they had a bad IP address in their /etc/hosts[16:47:31] *** kwinz3 has joined #opensolaris[16:47:38] <longcat> pkg.sun.com or pkg.opensolaris.org - same difference[16:47:44] <lblume> vraa: In any case, you'd rather look at smb.conf to force both access to the shares and permissions on the files created there.[16:48:08] <alanc> longcat: support is on pkg.sun.com, not pkg.os.o[16:48:13] <vraa> Stric: i do not understand permissions, i just want to eliminate them so i can access files[16:48:17] *** pizdec has joined #opensolaris[16:48:36] <longcat> alanc: try it out yourself http://pkg.sun.com/support[16:48:48] <rkeene> It became unavailable at the same time as Solaris 10 returned to requiring a license and Oracle announced that OpenSolaris support would be ending... I assumed that the /supported going away was a result[16:49:01] <lblume> vraa: You need to understand permissions if you're using any OS created in the last 20 years or so.[16:49:01] <vraa> yah i am using smb.conf, that's why i was so confused, i have set globally use SHARES and i had allowed users PER SHARE, like in [Home], but I will read more and find out why there was conflict. chmod is nice[16:49:06] <alanc> I don't know what's up there[16:49:11] <vraa> lol :( i've been using win98[16:49:52] <lblume> yes, MS-DOS is a bit older than that :-P[16:50:12] <betchou> vraa: so that's right you should have not any problem with 777 as rights ... but that's really not done for[16:50:51] <vraa> well from here, what i have to do is setup the proper permissions per group using the ACL inheritance codes[16:51:00] <betchou> vraa: usually youg get something like 755 for a directory, and 644 for a file[16:51:27] <betchou> vraa: you should google a bit on permissions[16:51:50] <tehhobbit> [16:53] < vraa> Stric: i do not understand permissions, i just want to eliminate them so i can access files[16:51:57] <vraa> i am, i am still trying to find the opensolaris man chmod, what are differences between the GNU[16:52:00] <tehhobbit> ooh sorry slip of mouse[16:52:05] *** sickness has joined #opensolaris[16:52:30] <betchou> vraa: don't think there is a so huge difference for this command[16:52:39] <|woody|> there is :)[16:52:56] <lblume> vraa: Using Samba, you'd better only set permissions only in smb.conf[16:53:00] <lblume> force them[16:53:08] <|woody|> you can't mod zfs ACL with the gnu chmod[16:53:10] <vraa> [09:44] <DerSaidin> vraa: man chmod and make sure your using the solaris one, not the gnu one[16:53:22] *** lattera has quit IRC[16:53:45] <vraa> lblume force them? i thought smb.conf was over ridden by whatever your osol thinks your files are in the first place[16:53:49] <betchou> |woody|: so there is at least the GNU part :p[16:53:51] *** lattera has joined #opensolaris[16:54:28] <betchou> vraa: ... this is an other problem ... you have to take at your manpath[16:54:53] <betchou> vraa: man chmod could redirect you to the GNU one or the solaris one[16:54:57] <lblume> vraa: No, smb.conf allows forcing owner, group, permissions. It's a long read, but a recommended one.[16:55:05] <vraa> what? i am using b134 osol[16:55:37] <vraa> perhaps i will return in a few hours, it will take me time to go through chmod man page, i just found it here http://docs.sun.com/app/docs/doc/819-2239/chmod-1[16:56:13] <betchou> vraa: try this one : echo $MANPATH[16:56:48] <vraa> betchou http://osol.pastebin.com/5FFvtFeR[16:57:26] <betchou> vraa: can you try in an non root account?[16:57:30] <vraa> http://osol.pastebin.com/1QJPRjFv <-- this is my smb.conf, it's very neat and organized[16:58:02] <vraa> sure betchou http://osol.pastebin.com/dzjRdwfi[16:58:08] <lblume> On osol, MANPATH is automatically derived frm the PATH by default[16:58:50] <betchou> vraa: so we should check your PATH instead ;)[16:59:08] <betchou> lblume: thanks :)[16:59:45] *** morettoni has joined #opensolaris[16:59:54] <vraa> can i log back in as "su -"[16:59:59] <vraa> or should i do it as "vraa"[17:00:03] *** morettoni has quit IRC[17:00:08] *** phimic has quit IRC[17:00:43] <vraa> http://osol.pastebin.com/5WcFHNYg < - i did it both watys[17:00:55] <vraa> *ways, so you can see i use the gnu thing i think[17:01:07] <betchou> vraa: exact :)[17:01:38] *** kwinz3 has quit IRC[17:01:50] <vraa> btw, how can i read my mail? on ubuntu i used to use "mutt" or "pine" but neither are available[17:03:06] <betchou> vraa: ... ... I think you can compile them :P or just install ... ;)[17:03:23] <kmays> vraa: they are...repositories like OpenCSW and Blastwave has those mutt/pine packages.[17:03:25] <jbit> just install SUNWmutt[17:03:31] <kmays> and IPS[17:03:32] <timsf> pkg:/mail/mutt[17:03:49] <vraa> a) compiling is ... outside my bounds right now, i tried to compile VLC and i had to reinstall osol[17:04:02] <jbit> vraa: pfexec pkg install SUNWmutt[17:04:03] <jbit> easy[17:04:06] <vraa> i thought all the SUNW stuff was renamed? that's what happened with the samba stuff[17:04:07] <betchou> http://forums.opensolaris.com/thread.jspa?threadID=193[17:04:08] <betchou> :P[17:04:10] <kmays> vraa: your ZFS ACLs handling in Samab are also done by patching or was.[17:04:17] <kmays> samba[17:04:18] <jbit> vraa: tehre are back links[17:04:33] <jbit> although timsf gave you the "new" name[17:04:52] <vraa> omg sunwmutt does exist, it's downloading and installing right now![17:05:02] <vraa> yeah i see mail/mutt, fantastic! i love this OS![17:05:10] <vraa> kmays: what do you mean by patching?[17:05:11] <longcat> usually my reaction when i see a package in ips[17:05:13] <betchou> vraa: ... that's a link for the other package[17:05:35] <betchou> vraa: some kind of trick for the renaming[17:06:01] *** hsp has quit IRC[17:06:06] <vraa> is it dangerous to add repositories? in ubuntu you could have mess-ups with multiple repo's[17:06:20] <kmays> look up Samba ZFS ACL patch[17:06:32] <tomww> vraa: vlc is best from a IPS repository or experimental/SFEvlc to compile by a RPM build recipe[17:06:38] <betchou> vraa: you should juste not try to install kde from Blastwave :P[17:06:39] <kmays> You'll have to tweak it for Samba 3.5.x.[17:06:53] <vraa> oh kmays maybe you are a few levels ahead of me[17:07:10] <betchou> vraa: you should only use other repository for packages without so much dependencies[17:07:16] <vraa> i am still studying the ACL model from zfsadmin.pdf[17:07:18] <kmays> betchou: :D[17:07:19] *** hsp has joined #opensolaris[17:08:01] *** nikolam_ is now known as nikolam[17:09:10] <lblume> There is no support for ZFS ACL in Samba. Sun proposed a patch, that was briefly integrated then rejected and removed (or never integrated).[17:10:25] *** tsoome has joined #opensolaris[17:10:49] <lblume> It was included in some Solaris Samba updates for a while, but it was still *very* buggy and not usable.[17:11:34] <vraa> well if it was buggy it's good they removed it[17:12:46] *** solar_sea has quit IRC[17:13:02] *** hunter has joined #opensolaris[17:13:51] <RoyK> hm... http://osol.pastebin.com/vQkxdePj <-- anyone got an idea? the iso file should be fine - I tried others as well[17:14:35] <vraa> *whoa* turns out "ls -V" used in GNU also doesn't show ZFS ACLs, use "/usr/bin/ls -V" and it'll show you more access control stuff[17:14:35] <lblume> I'd rather have had them making it work completely. But the problems lie deep in the core of Samba, where draft POSIX ACL are assumed, and can't be easily replaced by NFSv4 ones.[17:14:52] <nachox> i think there is still some support, with shadow_copy2[17:14:56] <lblume> Welcome to a brave GNU world.[17:15:02] <nachox> http://www.edplese.com/blog/2009/12/02/samba-shadow_copy2-enhancements/[17:15:33] *** thepickle has joined #opensolaris[17:15:50] <vraa> http://osol.pastebin.com/aALNQS76[17:16:36] <CosmicDJ> that's not a bug, that's a feature ;)[17:17:27] *** pegasus-rider has quit IRC[17:17:34] <lblume> and you should investigate why the group of your files is nobody, that's malsmost certainly wrong and not desirable.[17:18:08] *** merzo has quit IRC[17:22:20] <kmays> I agreed with Alan... Oracle said "support contract". They said that more details and refinement of the support contracts and roadmaps will come out by the time they released OSOL 2010.X.[17:22:35] *** Phoenixland has joined #opensolaris[17:22:41] <kmays> Pretty straightforward.[17:23:08] <Alasdairrr> As long as I can use OpenSolaris on Dell equipment and pay for updates, I'll be happy[17:23:28] <Alasdairrr> If I can't, then fuck Oracle, I'm off to Linux where I'll be unhappy but better off financially.[17:23:32] <Alasdairrr> *grumble*[17:23:33] <Alasdairrr> ;)[17:24:04] <lblume> It seems that's what Oracle wants anyway...[17:24:12] *** Phoenixland has quit IRC[17:24:31] *** Phoenixland has joined #opensolaris[17:24:31] <Alasdairrr> It does however look like Solaris 10/11 might be Sun/Oracle hardware only though[17:24:41] <lblume> Rich people used to throw money out the windows, not cheapskates like your average OpenSolaris ebthusiast....[17:24:44] <kmays> ;) - http://www.dell.com/content/topics/global.aspx/sitelets/solutions/management/solaris_solution?c=us&l=en&cs=555[17:24:45] *** lattera has quit IRC[17:24:56] <Alasdairrr> *clicks*[17:26:41] <Phoenixland> Is 2010.03 out?[17:26:50] <longcat> no.[17:26:52] <Alasdairrr> No - it'll be out in May/June[17:27:12] <Phoenixland> Ok tenks![17:28:02] <vraa> lblume: i removed the existing acl's by doing "/usr/bin/chmod -R A- /vkc_zfs_tank/[17:28:09] *** swift has joined #opensolaris[17:28:37] <vraa> wait nvm, i didn't do A= and A+ using -R, let me do -R[17:29:22] *** lattera has joined #opensolaris[17:29:56] <kmays> Alasdairrr: Oracle's priority support is of course their own hardware..so the OEM partners are the ones to go to if you want support for their 'custom' hardware..which then lies their own support contract. ;)[17:30:04] <vraa> http://osol.pastebin.com/bS7KT3R8 <-- there we go, now that looks better, i still have to fix the owner:group thing though[17:30:45] *** swift_ has quit IRC[17:30:48] <alanc> Alasdairrr: if it's important to you, be sure to let your Oracle sales reps know - they're getting lots of feedback from customers, and plans are always subject to change if they find out it loses more business than it gains[17:31:13] *** ewdafa has quit IRC[17:32:00] *** jthunder_ has quit IRC[17:32:22] <lblume> vraa: 777 looks rather evil to me. You should first fix your smb.conf to give access and force permissions.[17:32:24] *** Mace_N900 has joined #opensolaris[17:33:42] <lblume> Yes, it's such a smart way to do business, scare people away, then go "oh we were just kidding, please come back". It worked so well for Sun, they're now one of the greatest independant IT companies in the world. Oh, wait....[17:33:58] *** datadigger has quit IRC[17:34:04] <Spencer_tt> Alasdairrr: umm open source is always expensive as long as you don't know how to use it :D[17:34:08] <vraa> lblume: I agree, 777 means anyone can do anything right? let me figure this smb.conf out[17:34:09] *** betchou has quit IRC[17:34:45] <vraa> lblume: when you say give access and force permissions, what does that mean?[17:35:45] *** stevel has joined #opensolaris[17:35:54] *** thepickle has quit IRC[17:35:54] *** biant92 has quit IRC[17:35:54] *** astra-x has quit IRC[17:35:54] *** nikolam has quit IRC[17:35:54] *** xtrondo has quit IRC[17:35:54] *** Triskelios has quit IRC[17:35:54] *** moepmoep has quit IRC[17:35:54] *** Edgeman has quit IRC[17:35:54] *** Tusk has quit IRC[17:35:54] *** lisppaste3 has quit IRC[17:35:54] *** dunc has quit IRC[17:35:54] *** Berny has quit IRC[17:35:54] *** IvanR__ has quit IRC[17:35:54] *** mmealling1 has quit IRC[17:35:54] *** pjama has quit IRC[17:35:54] *** Tempt has quit IRC[17:35:54] *** frankS2 has quit IRC[17:35:54] *** evalenci has quit IRC[17:36:01] *** ewdafa has joined #opensolaris[17:36:05] <alanc> lblume: yep, Oracle knows not to make some of Sun's mistakes like "Give everything away for free and hope the money fairy leaves a few billion under your pillow one night", but will probably have to re-learn a lot more the hard way[17:36:26] <lblume> vraa: Basically, look at "force create" in smb.conf(4)[17:36:59] *** thepickle has joined #opensolaris[17:36:59] *** biant92 has joined #opensolaris[17:36:59] *** astra-x has joined #opensolaris[17:36:59] *** nikolam has joined #opensolaris[17:36:59] *** xtrondo has joined #opensolaris[17:36:59] *** Triskelios has joined #opensolaris[17:36:59] *** moepmoep has joined #opensolaris[17:36:59] *** Edgeman has joined #opensolaris[17:36:59] *** Tusk has joined #opensolaris[17:36:59] *** lisppaste3 has joined #opensolaris[17:36:59] *** dunc has joined #opensolaris[17:36:59] *** frankS2 has joined #opensolaris[17:36:59] *** Berny has joined #opensolaris[17:36:59] *** IvanR__ has joined #opensolaris[17:36:59] *** mmealling1 has joined #opensolaris[17:36:59] *** pjama has joined #opensolaris[17:36:59] *** Tempt has joined #opensolaris[17:36:59] *** evalenci has joined #opensolaris[17:36:59] *** pratchett.freenode.net sets mode: +o Tempt[17:37:04] *** twisti has quit IRC[17:37:21] *** niq has joined #opensolaris[17:37:21] *** niq has joined #opensolaris[17:37:24] <Spencer_tt> Oracle can learn from Ubuntu or Redhat/Fedora on giving away discs?[17:38:18] *** e1kg has joined #opensolaris[17:39:08] <lblume> alanc: Well, the Redmond fairy did bring some money, right? But Sun's problem was that they never were never trying to sell their products, and barely knew how to receive money.[17:39:14] <vraa> lblume: i see there is more than force create, i'll have to use force user and force group right?[17:39:50] <kmays> Alasdairrr: Also, OpenSolaris is 'conceptual' so the OEMs don't see it as an actual 'product' like they do Oracle Solaris...so the target product on the table is really the "Oracle Solaris" product - in which the future release is based on 'OpenSolaris + Value Add' technologies. The OSOL binary releases have features that can be added/dropped at any predetermined time meaning they might or...[17:39:52] <kmays> ...might not be in the final product release (so they say). Yet, the model is pretty much the Fedora=>RHES type model for simplistic debating.[17:40:09] <alanc> I've got free discs from Oracle before - they've done it before, so presumably know when it's worthwhile[17:40:36] <alanc> I think they even offer free cd's for Oracle Enterprise Linux, though that's a somewhat different market[17:40:36] <vraa> so that means i can create a user on osol machine called "sambauser", force user = sambauser, assign all files to be owned by sambauser, however the users still login to authenticate using their regular login (right now only vraa and root being the possible users)[17:40:42] <lblume> vraa: Not necessarily if there's no constraint on the unix side.[17:41:32] <lblume> Well, Oracle was putting their DB CD in Solaris' boxes for years, didn't they?[17:42:17] <Spencer_tt> well Ubuntu managed to get well popular thanks to it's free cd program, to people who are interested it usually pays of one way or the other[17:42:20] <alanc> yeah, I think it came with a bunch of OS releases[17:42:45] *** Fullmoon has joined #opensolaris[17:43:13] <Spencer_tt> it would be quite a bundle to get OpenSolaris + Oracle on one disc or set of discs[17:43:20] *** Fullmoon has quit IRC[17:43:47] *** swift has quit IRC[17:44:13] *** swift has joined #opensolaris[17:44:57] <kmays> Alasdairrr: Most ISVs/IHVs follow that theory in sticking with the actual 'product' releases or the officially supported and tested OpenSolaris binary releases (ie. like OSOL 2008.05 or 2009.06).[17:46:21] <lblume> Are there *any* ISV-supported application for any osol release?[17:47:09] <|woody|> flluendo codecs :)[17:47:38] <kmays> A few ISVs started doing certifications when OSOL 2008.05 was released.[17:48:11] <kmays> Same with IHVs.[17:48:15] <kmays> Like Nvidia.[17:48:51] <lblume> kmays: If they started on 2008.05, they must have run away so fast, they must be on the other side of the moon by now :-)[17:49:10] <|woody|> It's just lot of man power to put in testing twice a year[17:49:20] <|woody|> so I guess most will stick to solaris[17:49:38] <lblume> Yes, so my question is still unanswered, are there any applications certified for sol now?[17:49:42] <lblume> osol[17:49:44] <mui> I'd prefer dedup fix over release[17:49:49] <mui> wonder whats status with it[17:50:33] *** baza11215 has joined #opensolaris[17:50:46] *** pablomh has joined #opensolaris[17:50:49] *** jthunder_ has joined #opensolaris[17:51:18] *** Kementari has joined #opensolaris[17:52:06] <Spencer_tt> lblume: good question I bet time will answer that question[17:52:56] <lblume> Well, a simpler one then: are then *any* ISV certified applications for Fedora?[17:53:39] <lblume> Since time and again, we get that RHEL/Fedora comparison bullshit.[17:54:24] <vraa> fedora doesn't have zfs though, so why would anyone use it[17:54:44] <tsoome> no linux has zfs..... so.. .P[17:54:50] <Spencer_tt> yes novel has for suse and redhat has it's own[17:54:56] *** merzo has joined #opensolaris[17:55:27] * lblume should look into that btrfs thing some day.[17:55:31] <Spencer_tt> what's the point behind isv certification and does sun have it;s own program[17:55:57] <vraa> tsoome :D :D[17:56:06] *** IPGHOST has joined #opensolaris[17:56:41] <Alasdairrr> mui: "dedup fix" - what's bust with it? Hope i'm not about to suffer some catastrophic data loss /o\[17:56:56] <tsoome> Spencer_tt: well, how i know i can use vxvm with vxfs?[17:57:52] <Spencer_tt> maybe it's only a matter of time one winux has zfs...[17:57:54] *** biant92 has quit IRC[17:57:59] <Spencer_tt> tsoome: you tell me :)[17:58:07] *** mroconnor has joined #opensolaris[17:58:14] <tsoome> they cant have it;)[17:58:27] *** biant92 has joined #opensolaris[17:58:56] <Spencer_tt> compatible vxwm & vxfs programs :p ?[17:59:04] <tsoome> no, zfs[17:59:15] <Alasdairrr> Linux are rapidly developing brtfs. Some nice features, but it seems a rushjob.[17:59:19] <Spencer_tt> I bet it's scary :D[17:59:37] <vraa> Alasdairrr what are some indicators that it's a rush job?[18:00:11] <Alasdairrr> http://lwn.net/Articles/358940/[18:00:50] <kmays> lblume: Actually, Sun did have a list of products that "claimed to work" on OpenSolaris and had some docs on it...let me see if they've been scrubbed as I had some in my archives.[18:01:56] <lblume> kmays: "Claimed to work", as in "fully compatible and supported"?[18:02:00] <Shoggoth> is it possible to download ips packages for offline installation?[18:02:02] *** houst0n has left #opensolaris[18:02:27] *** thevraa has joined #opensolaris[18:03:34] <tsoome> no.[18:03:35] *** twisti has joined #opensolaris[18:03:38] *** vraa has quit IRC[18:04:00] <Shoggoth> that's inconvenient to put it politely[18:04:22] <melbogia> Is it possible to change the number of auto-snapshots kept on a per dataset basis?[18:04:30] *** IPGHOST has quit IRC[18:04:34] *** Erwann has quit IRC[18:04:57] *** perlmongo has joined #opensolaris[18:05:35] <timsf> melbogia: nope, but you can create additional service instances and have those instances only snapshot given datasets.[18:05:56] <melbogia> oh interesting idea[18:05:56] <thevraa> Alasdairrr that is a old article isn't it?[18:05:58] <kmays> lblume: Well, compatible ..that support thing is another subject that required a higher certification status more IHV/ISV related and passing Sun's Level 1/2 tests. I'm sure you've seen this on their website. But the OEM/IHV/ISV are the ones that can actually slap the 'we support our products on this OS release' statement moreso than Sun.[18:06:31] <kmays> I mean Oracle.[18:07:01] *** Shoggoth has quit IRC[18:07:08] <kmays> Let me find that old doc.[18:08:06] *** fleers has quit IRC[18:08:42] *** Shoggoth has joined #opensolaris[18:09:04] <lblume> Sure, I'd be interested in even a single example of a real production app.[18:09:17] <kmays> I think a lot of that certification died awhile ago though.[18:09:44] <lblume> Was there ever any?[18:11:35] <tsoome> i know there have been references to certified apps, but myself have never read what exactly such certification will prove[18:11:46] <kmays> It started back during the OSOL 2008.05 release but then never rocketed to stardom.[18:12:27] <lblume> I want a name, not vague hearsay.[18:12:45] <tsoome> man google:P[18:12:49] <tsoome> :D[18:12:59] <kmays> tsoome: only that it worked. That was it pretty much.[18:13:03] *** dijenerate has quit IRC[18:13:44] <kmays> Kinda like the Flash/Adobe Reader era on OpenSolaris.[18:13:49] *** milek has quit IRC[18:13:52] <RoyK> anyone here using xvm?[18:14:05] <lblume> tsoome: Whatever, I was not the one pretending there were.[18:14:35] *** sergiusens has joined #opensolaris[18:15:42] <Triskelios> kmays: "era"?[18:15:43] <lblume> Even Sun never gave any serious support guarantee for stable osol release, just a "best effort" one.[18:16:17] <kmays> One note: tsoome - it really goes to saying whom is certifying on Solaris 10 today (or in the past 2-3 years)?[18:16:38] *** merzo has quit IRC[18:17:36] <lblume> kmays: You're a politician, right? Never answer with "Yes/No/I don't know", but find some vaguely tangent topic and go for it?[18:17:57] <kmays> lblume: Oh, you mean like Nvidia, Toshiba, Dell, others?[18:18:24] <kmays> I didn't think I have to spell it out...[18:18:59] *** kwinz3 has joined #opensolaris[18:19:14] <kmays> I did say OEM/ISV/IHV make those claims..not necessarily Sun/Oracle.[18:19:21] <lblume> nvidia, Dell, Toshiba are hardware makers, not ISV.[18:19:39] <kmays> But they did publish a doc a log time ago.[18:19:45] <kmays> Umm...drivers...[18:19:53] <lblume> Well, yes, before you went for the tangent IHV topic, I had asked speciifically about ISV.[18:20:15] <lblume> Yes, drivers are not ISV applications[18:20:25] <lblume> Oh well, I'm wasting time, you just don't know.[18:20:28] <stallion_work> gdamore: ping[18:21:02] <kmays> No..you said ISV.... apps is specific and again I mentions apps compatible with OpenSolaris.[18:21:20] <kmays> I said support is another topic.[18:21:24] *** afrasi has joined #opensolaris[18:22:26] <lblume> You did not even mention an app name. "Toshiba" is not an app.[18:22:30] <lblume> Aaaargh, biting again![18:22:34] *** ziyourenxiang has quit IRC[18:22:36] <lblume> I'm out for the night, bye all.[18:22:43] <kmays> Oh..but someone mention to just query the Internet.[18:22:50] *** lblume has quit IRC[18:22:51] <kmays> I could say bacula![18:23:07] <kmays> I could mention a bunch of people..but then we'd wan tto talk about support.[18:23:30] <gdamore> stallion_work: pong[18:23:32] *** derchris has quit IRC[18:23:35] <kmays> and what release of OSOL specifically.[18:23:37] *** derchris has joined #opensolaris[18:23:45] <stallion_work> gdamore: quick question - looks like efe will be done by this weekend[18:23:49] *** twisti has quit IRC[18:23:58] <gdamore> cool. renamed to spwr?[18:24:02] <stallion_work> the device is ridiculously simple - even simpler than afe.[18:24:09] <gdamore> yep.[18:24:14] <stallion_work> meh. probably.[18:24:20] <gdamore> no model variants to worry about, unlike afe.[18:24:29] <stallion_work> I'll keep it as efe in emancipation, but for the integration, I'll do a mass rename[18:24:48] <stallion_work> I still think efe is superior :D[18:24:53] *** twisti has joined #opensolaris[18:25:07] <gdamore> ok. so we'll need a PSARC automatic review case, and I'll need to get NICDRV results from you.[18:25:13] *** InTheWings has joined #opensolaris[18:25:14] <stallion_work> *nod*[18:25:25] <stallion_work> I'm hoping I can get NICDRV done while I'm down in Athens.[18:25:32] <stallion_work> so my goal is to be code complete by Sunday[18:25:35] <gdamore> i don't disagree that efe is a nicer name. but spwr is already in use, and it will be less painful for people converting.[18:25:41] <stallion_work> true[18:25:44] <gdamore> cool.[18:25:55] *** alfism has joined #opensolaris[18:26:15] <stallion_work> Anyhow, the device tends to play mommy when it comes to bad frames[18:26:20] *** dijenerate has joined #opensolaris[18:26:38] <stallion_work> I'm wondering if its worthwhile from a statistical perspective to accept run frames and/or other bad bits[18:26:39] <gdamore> ?[18:26:52] <stallion_work> basically you get counters for alignment and crc[18:26:54] <gdamore> they should be counted.[18:26:58] <stallion_work> nothing for runt frames[18:27:03] <stallion_work> so I'd have to set the filter to accept them[18:27:12] <kmays> (as a last comment: a company like Intel claims "OpenSolaris certified on over 1000 Intel-based systems")[18:27:22] <gdamore> i'd go ahead and accept 'em, so you can count 'em. (drop em during rx.)[18:27:27] <stallion_work> the only other option would be to use receive lookahead which would make things more complex[18:27:29] <stallion_work> okay[18:27:45] *** vraa has joined #opensolaris[18:27:50] *** sergiusens has quit IRC[18:27:54] <gdamore> there should not be many runts on the wire... if there are its a site problem.[18:27:58] <vraa> hi, everyone, i found a great link which helps explain why one should use samba instead of kernel cifs: http://jimcollier.blogspot.com/2010/03/2kc83jor.html#idNetworkSharing[18:27:59] <stallion_work> *nod*[18:28:12] <stallion_work> I could look into lookahead as well - that would make gathering stats quite a bit easier[18:28:12] <gdamore> (in fact, I've never seen a runt on a properly configured network.)[18:28:16] <stallion_work> but it complicates the rx path[18:28:29] <stallion_work> most devices will do the proper padding[18:28:43] <stallion_work> but some of the older ones like re will silently send bad frames[18:28:46] *** thevraa has quit IRC[18:28:50] *** chrismat2 has joined #opensolaris[18:28:52] <gdamore> i wouldn't waste time with lookahead... not worth the added risk.[18:28:57] <stallion_work> *nod*[18:29:05] *** chrismat1 has quit IRC[18:29:06] <stallion_work> right now its at a smidge under 1.2KLOC[18:29:13] <gdamore> nice.[18:29:18] <stallion_work> rx is almost complete - and tx won't take much[18:29:26] <gdamore> haven't cracked the 1K barrier though.[18:29:27] <stallion_work> so I'm thinking I'll definitely be below 1.5KLOC[18:29:33] *** Mace_N900 has quit IRC[18:29:36] <stallion_work> not yet :)[18:29:41] <stallion_work> but I tend to use a lot of veritcal space[18:30:17] <gdamore> heh.[18:30:20] <gdamore> I do too.[18:30:30] <gdamore> afe is 2.4KLOC. mxfe is 3KLOC.[18:31:14] *** scottdickson has quit IRC[18:31:19] *** chrismat2 has quit IRC[18:31:35] <jbk> vraa: i think that's not entirely accurate, though part of it is due to project indiana[18:31:49] *** psychicist has quit IRC[18:31:54] <jbk> AFAIK, if any of the actual posix tools are munging up attributes or ACLs, it's a bug[18:32:12] <jbk> the problem is, none of the GNU utilities support them[18:32:16] <vraa> i see, thx jbk, i have class now but i'll be back in a few hours to do more research[18:32:20] <jbk> which are first in the path (by default)[18:32:21] <RoyK> stallion_work: what is KLOC?[18:32:23] <vraa> my goal is to create a zfs file server :)[18:32:40] *** hunter has quit IRC[18:32:50] <jbk> and yes, the cmdline syntax is a bit annoying (though I don't know of any better cmdline stuff)[18:33:28] <jbk> the aclinherit and aclmode defaults are kinda crappy, but i've found things work much better when you change them to sane values[18:33:53] <jbk> and to say that opensolaris won't support smb2 i think is quite dumb[18:34:01] <Stric> RoyK: kilo lines of code[18:34:13] <jbk> given that the in-kernel cifs server forms the basis of the CIFS support for their storage appliance[18:34:23] <jbk> that would imply that oracle would never update it to support the new protocol[18:34:40] <jbk> which seems unlikely[18:37:09] *** psychicist has joined #opensolaris[18:37:14] *** albakry has joined #opensolaris[18:37:37] *** fleers has joined #opensolaris[18:38:14] *** melbogia has quit IRC[18:38:41] *** MuffiXXL has joined #opensolaris[18:39:37] * stallion_work &[18:41:29] <Hedonista> i have borked some file permissions on a pool i share locally at my home network, i am able to mount the share but cannot see or do anything , how can i restore the original permissions or should i try making a new pool and importing the files to that?[18:42:39] <CosmicDJ> nfs/smb?[18:42:49] <Hedonista> nfs CosmicDJ[18:43:10] <Netwolf> hi folks. I'm trying to install Opensolaris or SXCE build 31 on my blade 2500 which has a xvr-1200. alanc here was kind enough to tell me that that opensolaris only supports xvr-100 in the pci line. However, I found this link http://hub.opensolaris.org/bin/view/downloads/sparc_graphics which seems to contain drivers. Can or would these drivers simply work by installing them in Opensolaris or are they just for refrence purposes[18:43:11] <CosmicDJ> v3 or v4?[18:43:29] <Hedonista> v3[18:43:44] <CosmicDJ> your user id's match? i.e. it's the same on the client *and* on the server?[18:44:38] <Hedonista> i set max nfs version to 3 on server to match client[18:44:42] *** dunc has quit IRC[18:45:09] <CosmicDJ> http://en.wikipedia.org/wiki/User_identifier_(Unix)[18:45:21] <Hedonista> CosmicDJ, thanks[18:47:27] *** nikolam has quit IRC[18:48:18] *** nikolam has joined #opensolaris[18:49:28] *** twisti has quit IRC[18:49:42] *** twisti has joined #opensolaris[18:50:24] *** Fullmoon has joined #opensolaris[18:51:41] *** yippi has joined #opensolaris[18:53:30] *** kwinz3 has quit IRC[18:55:55] <CIA-21> Johnny Cheung <Johnny.Cheung at Sun dot COM>: 6943734 parallel rdmacm operations on same CMID can cause panic with rdsv3, 6943145 solaris user verbs api change to include IBV_ACCESS_SO flag and functionality, 6944892 ibt_ip_path_handler_t(9e) is passed src_ip_p > required, 6945146 sol_ucma: codes cleanup for issue reported by coverity tools, 6945163 sol_uverbs: codes cleanup for issues reported by coverity tools[18:57:47] *** melbogia has joined #opensolaris[19:01:17] *** panos has joined #opensolaris[19:01:25] *** KermitTheFragger has quit IRC[19:01:59] *** Phoenixland has quit IRC[19:02:14] *** Mace_N900 has joined #opensolaris[19:02:26] *** nitrile has quit IRC[19:02:56] *** Phoenixland has joined #opensolaris[19:02:56] *** xtrondo has quit IRC[19:03:09] *** xtrondo has joined #opensolaris[19:05:15] *** albakry has quit IRC[19:05:41] *** albakry has joined #opensolaris[19:06:03] *** sipior has quit IRC[19:06:08] *** forquare has quit IRC[19:06:22] *** blizzow has joined #opensolaris[19:07:25] *** jgracin has quit IRC[19:08:27] *** forquare has joined #opensolaris[19:11:02] *** vraa has quit IRC[19:13:35] *** bubbva has joined #opensolaris[19:14:35] *** kwinz3 has joined #opensolaris[19:14:40] *** nevdull has joined #opensolaris[19:14:54] <Shoggoth> <flame>Who the hell decided it was a good idea not to allow offline installation of IPS packages</flame>[19:15:06] *** sergiusens has joined #opensolaris[19:16:04] *** nevdull has quit IRC[19:16:14] <Alasdairrr> you can fire up your own IPS server fairly easily[19:16:22] <longcat> at least now isos of the repo are distributed[19:16:24] <Shoggoth> I don't want to download the whole lot[19:16:31] *** albakry has quit IRC[19:16:32] <Shoggoth> I have limited bandwidth and quota[19:16:46] <Alasdairrr> Well all the software on the CD gets installed[19:16:53] <Alasdairrr> So if you want additional software, it has to come from "somewhere" :)[19:17:08] <Alasdairrr> And limited bandwidth sounds like a personal problem ;-P[19:17:14] <longcat> you can pkgrecv on a machine with internet access onto a repo that your internal machines can use[19:17:34] <Shoggoth> yes... but I just tried installing sunstudio and it bombed out at 297MB... it's 315MB in size... no I have to start over[19:17:42] <Shoggoth> I can't afford the wastage[19:17:43] <longcat> pkgrecv -s http://pkg.opensolaris.org/release -d http://localhost/myrepo some_fmri@version[19:17:50] *** vraa has joined #opensolaris[19:18:01] <longcat> probably requires other stuff. ask the guys in #pkg5[19:18:13] <Shoggoth> and if it bombs out half way through does it cache or do I have to restart[19:19:09] <Shoggoth> Alasdairrr: limited bandwidth is a personal problem for everyone in this country[19:19:36] *** hsp has quit IRC[19:20:29] <Alasdairrr> Would an IPS Mirror in Australia help?[19:20:35] <Shoggoth> nope[19:20:43] <Shoggoth> the problem is the ISP's here[19:20:50] <Shoggoth> they charge like the proverbial wounded bull[19:20:57] <Alasdairrr> Give it 10 to 20 years[19:21:00] <Shoggoth> lol[19:21:01] <Shoggoth> yep[19:21:03] <Alasdairrr> :-)[19:21:13] *** albakry has joined #opensolaris[19:21:15] <Spencer_tt> Alasdairrr: limited bandwidth or limited patience or methods :D[19:21:23] <Alasdairrr> in the mean time perhaps you can get someone to post you the DVD of the IPS repo[19:21:31] <Spencer_tt> for other people that sounds like exploits..[19:21:43] <Spencer_tt> don't tell anyone :p[19:21:53] *** Phoenixland has quit IRC[19:21:57] <Alasdairrr> Sounds like I should move to .au and start an ISP[19:21:58] <Alasdairrr> :p[19:22:04] <Spencer_tt> :D[19:22:19] <Shoggoth> that wouldn't be much fun[19:22:31] <tsoome> why?[19:22:35] <Spencer_tt> don't forget other parts of the world you might become a millionaire in dollars :D[19:22:38] <Shoggoth> we have an incumbant monopoly telco that screws everyone over[19:22:57] <Alasdairrr> In the UK that used to be BT[19:22:59] *** Phoenixland has joined #opensolaris[19:23:02] *** ivo_ has joined #opensolaris[19:23:10] <Alasdairrr> But thankfully Ofcom came along and started beating them with a stick[19:23:17] <Alasdairrr> (Regulator)[19:23:34] <Alasdairrr> Things are "reasonably okay" now all things considered[19:23:40] <Shoggoth> yeah, well our regulator doesn't have as many teeth and our population is somewhat more spread out than yours[19:23:48] <Alasdairrr> *nods*[19:23:55] *** fOB has joined #opensolaris[19:23:56] *** swift has quit IRC[19:23:57] <Spencer_tt> don't worry someone can always beat ofcom with a bigger stick - that's why life is kinda interesting[19:25:47] *** hsp has joined #opensolaris[19:25:57] *** afrasi has quit IRC[19:26:09] <Shoggoth> I still don't understand why the damn package installer insists on downloading all 12 required packages again when it bombed out on the last one[19:26:27] <Alasdairrr> immaturity[19:26:34] <Alasdairrr> You can probably file a bug[19:26:38] <Alasdairrr> defect.opensolaris.org[19:26:49] <Alasdairrr> Even if there's another bug, the duplicates will piss off the right people ;-)[19:26:57] <Alasdairrr> although it's probably polite to search first[19:27:10] *** merzo has joined #opensolaris[19:27:18] <Shoggoth> lol[19:27:22] <Shoggoth> fair cop[19:27:48] <Alasdairrr> IPS is fairly young and still has a long way to go.[19:28:03] <Alasdairrr> Does anyone know if OpenSolaris will ever support sparse zones again? I had an idea for implementing them.[19:28:04] *** hsp has quit IRC[19:28:24] <Alasdairrr> Which was to have an ipkg install base that the global zone manages that all the non-global zones share[19:28:32] *** hsp has joined #opensolaris[19:28:57] <Shoggoth> lol... turns out our fascist government is good for something afterall... seems most of the packages were in the great-firewall-of-australia's cache[19:29:13] <Alasdairrr> you do have a fascist government[19:29:14] <Alasdairrr> what's with that?[19:29:20] <Alasdairrr> can't you uprise and shoot them in the head?[19:29:26] <Shoggoth> pretty sure I'll get billed for both[19:29:31] <Shoggoth> tempting[19:29:45] <Tonnerre> Alasdairrr, they tried that but it turned out to only weaken the mental capabilities of the government[19:30:02] *** hunter has joined #opensolaris[19:30:12] <Shoggoth> you haven't seen the news items on 'el reg and elsewhere about our compulsory firewalling?[19:31:00] <Alasdairrr> Yes I believe I have[19:31:00] *** baza11215 has quit IRC[19:31:17] *** baza11215 has joined #opensolaris[19:31:30] <Shoggoth> unfortunately the opposition are more fascist if that's possible[19:31:32] <Shoggoth> :-|[19:31:49] *** baza11215_ has joined #opensolaris[19:31:53] <Shoggoth> looks like you've guys have got yourself a three legged race....[19:32:14] <Alasdairrr> So it would seem[19:32:23] <Alasdairrr> Might even lead to electoral reform[19:32:26] <Alasdairrr> which would be nice![19:32:37] <Alasdairrr> It would mean my votes for the greens and the lib dems would finally count[19:32:46] <Shoggoth> can you please export some of that here....[19:32:49] <Shoggoth> pretty please[19:32:58] <Alasdairrr> I'll get right on that :-P[19:33:03] <Shoggoth> :)[19:34:30] <Alasdairrr> bugger, 6:40pm and I still have to cycle home[19:34:44] <tsoome> stop whining, you have no clue what is fascism or communism:P[19:35:13] *** jthunder_ has quit IRC[19:35:26] *** n1c0b has joined #opensolaris[19:35:43] *** n1c0b has left #opensolaris[19:36:03] *** baza11215 has quit IRC[19:36:03] *** baza11215_ is now known as baza11215[19:36:59] <Shoggoth> maybe not... but I'm pretty sure neither potential government party here are communist :)[19:37:14] *** cybernd has joined #opensolaris[19:37:47] *** hsp has quit IRC[19:40:19] *** Odin- has joined #opensolaris[19:40:52] *** Animal-X_ has joined #opensolaris[19:43:49] <kmays> I was referring to this ISV program earlier: http://wikis.sun.com/display/OSOLAppsContest/OpenSolaris+Application+Competition;jsessionid=05281FDA95701AC67155FEED0B7A6599[19:44:23] <kmays> Which had a few ISVs which passed...but this information is now elsewhere.[19:44:33] *** Animal-X has quit IRC[19:45:03] <kmays> Sun Software Library seems defunct.[19:45:34] *** ultrasparc-viii has joined #opensolaris[19:45:37] *** sndcrb has joined #opensolaris[19:45:41] *** vraa_ has joined #opensolaris[19:45:51] *** Kementari has left #opensolaris[19:47:19] <vraa_> when someone stops using acl, and goes back to the old method, does that mean i can use the gnu version of chmod now?[19:48:10] <kmays> But the new site is at: http://solutions.oracle.com/[19:48:34] *** hsp has joined #opensolaris[19:48:34] *** comay has joined #opensolaris[19:49:34] *** ChanServ sets mode: +o comay[19:50:09] <tsoome> why you need gnu chmod at all?![19:51:59] <vraa_> well i was using /usr/bin/chmod -V to display the permissions because gnu chmod doesn't[19:52:23] <vraa_> however, now i've read it's not good to use zfs acl on samba, because you have to forceuser and forcegroup and force create masks and stuff[19:52:50] <tsoome> i did ask why you need gnu chmod, not why you used solaris chmod...[19:53:05] <vraa_> so i did the osol chmod -R A- on all the files, and am starting from chmod 777, and working backwards as i read through man chmod[19:53:20] *** jiraia has joined #opensolaris[19:53:49] *** twisti has quit IRC[19:54:03] <vraa_> oops was i not supposed to do something?[19:54:36] *** Animal-X_ is now known as Animal-X[19:54:44] <CIA-21> Alexandre Chartre <Alexandre.Chartre at oracle dot com>: 6939707 domain can panic on 'ldm stop'[19:54:44] <CIA-21> Peter Shoults <Peter.Shoults at Sun dot COM>: 6885914 KDC doesn't enforce the password min-age policy[19:54:44] <CIA-21> Ali Bahrami <Ali.Bahrami at Oracle dot COM>: 6932220 ld -z allextract skips objects that lack global symbols[19:54:52] <Triskelios> vraa: ACLs work fine with the CIFS server, but Samba isn't aware of them[19:55:06] <tsoome> no, just curious. i just fail to see any benefit from gnu chmod 644 versus solaris chmod 644[19:55:33] *** twisti has joined #opensolaris[19:56:45] *** rcarcamo has joined #opensolaris[19:56:47] <tsoome> well, obviously its rather question about sanity of oracle developers including that gnu directory at all;)[19:57:50] <vraa_> well actually i was using osol chmod by typing it out "/usr/bin/chmod -V"[19:58:03] <vraa_> so gnu chmod is normal "chmod -R 777 /directory"[19:58:10] <tsoome> if i recall, samba should support zfs acl's[19:58:35] <tsoome> you dont need gnu chmod to do chmod -R 777[19:58:57] <tsoome> ;)[19:59:22] <Triskelios> ah, yeah, ZFS/NFSv4 ACL support was added to Samba[19:59:32] <vraa_> okay cool that's easy, these ACLs are easy, "/usr/bin/chmod -R -A=everyone@:full_set:fd:allow /directory"[19:59:43] <vraa_> yeah but i heard it was removed because it was buggy[20:00:12] <vraa_> tsoome, but gnu chmod doesn't require me to type "/usr/bin/chmod"[20:00:21] <tsoome> fix your path:P[20:00:47] <vraa_> i am not there yet[20:00:56] <vraa_> i dont do customizing because i reformat a lot[20:00:59] <tsoome> vi .profile :P[20:01:08] <tsoome> :D[20:01:41] <vraa_> Triskelios, do you know if it was added after it was removed?[20:02:16] <Triskelios> vraa_: say again?[20:02:50] *** raichoo has joined #opensolaris[20:02:57] <vraa_> it was said that acl support was removed from samba because it was buggy[20:03:05] *** sergiusens has quit IRC[20:03:42] <Triskelios> http://src.opensolaris.org/source/xref/sfw/usr/src/cmd/samba/Makefile.sfw shows "--with-shared-modules=vfs_zfsacl"[20:04:31] <vraa_> okay i will show that to lblumel i think[20:05:20] <vraa_> so i can continue adding the ACL method of chmod'ding ya?[20:07:21] *** Phoenixland has quit IRC[20:07:27] *** vraa__ has joined #opensolaris[20:07:38] *** fOB has quit IRC[20:07:40] *** Phoenixland has joined #opensolaris[20:08:14] *** albakry has quit IRC[20:08:47] <Triskelios> vraa: if the ACL support works, then sure[20:09:08] <vraa__> it's very easy to do ACL[20:10:01] *** vraa_ has quit IRC[20:11:41] *** xtrondo has quit IRC[20:12:14] *** Odin- has quit IRC[20:12:55] *** Odin- has joined #opensolaris[20:13:35] *** Shoggoth has quit IRC[20:13:54] *** fleers has quit IRC[20:15:51] *** syamajala has joined #opensolaris[20:16:47] *** MuffiXXL_ has joined #opensolaris[20:18:46] *** forquare has quit IRC[20:19:06] *** MuffiXXL has quit IRC[20:19:07] *** MuffiXXL_ is now known as MuffiXXL[20:19:43] *** forquare has joined #opensolaris[20:21:12] <vraa__> Triskelios, what version source were you referencing?[20:21:17] <vraa__> 3.0.37 samba?[20:22:36] *** dunc has joined #opensolaris[20:22:52] <Triskelios> seems to be[20:24:02] <vraa__> okay well i will try "security = domain"[20:24:23] <vraa__> maybe having the win2k8 ad ds handle user auth will fix this issues im having[20:25:20] <stallion_work> gdamore: mii question if you have a moment?[20:25:44] <high-rez> I've been thinking of using opensolaris to export zvols as iscsi targets (sun x4500) to a machine running vmware esxi (sun x4600)... Does this sound reasonable?[20:26:02] <Triskelios> vraa: is idmap running?[20:26:37] <Triskelios> high-rez: it's a common use of the iSCSI support[20:27:01] *** datadigger has joined #opensolaris[20:27:28] <high-rez> Triskelios: Any gotchas I should know about? I tried this with solaris 10 that came pre-isntalled by esxi would discoerable but didn't like the disks for some reason (not sure if I did something wrong in creating them)[20:28:00] <high-rez> (i unded up deleting the iscsi service in the process without really knowing what I was doing : -)[20:30:20] <vraa__> no i turned off idmap, why do i need it?[20:30:22] *** milek has joined #opensolaris[20:31:56] <Triskelios> vraa__: looks like you don't, so ignore me[20:32:15] <vraa__> well i turned it off because i thought it was part of kernel cifs[20:33:26] *** timeless_mbp has quit IRC[20:34:16] *** albakry has joined #opensolaris[20:34:45] *** timeless_mbp has joined #opensolaris[20:34:46] *** timeless_mbp has joined #opensolaris[20:35:53] *** Mousk4r1 has quit IRC[20:36:00] *** fleers has joined #opensolaris[20:44:28] *** hunter has quit IRC[20:47:21] *** hunter has joined #opensolaris[20:47:43] <ultrasparc-viii> hey all, I just installed mysql51 on opensolaris using "pkg install SUNWmysql51" and for some reason the app isn't showing up when I run "svcs" ... am I missing a step here?[20:48:10] <trygvis> on opensolaris you have to register your own manifests[20:48:47] <ultrasparc-viii> ok that's what I figured. how do I do that[20:49:01] <trygvis> svcadm import <file> iirc[20:49:52] <bdha> svccfg[20:50:17] *** jthunder_ has joined #opensolaris[20:51:22] <ultrasparc-viii> svccfg import /var/svc/manifest/application/database/mysql_51.xml[20:51:23] <ultrasparc-viii> :)[20:51:27] <ultrasparc-viii> there we go![20:52:31] *** Aria has joined #opensolaris[20:54:13] <longcat> careful, 51 has some bugs that will make it crash[20:54:16] <Triskelios> trygvis: you don't have to, but pkg doesn't import them so you'd have to either restart manifest-import manually or reboot[20:54:27] <longcat> unless you use 64bit 51[20:54:34] <trygvis> right[20:55:17] *** eleusis has quit IRC[20:56:19] *** Hedonista has quit IRC[20:56:45] *** eleusis has joined #opensolaris[20:56:45] *** eleusis has joined #opensolaris[20:58:18] *** cslamar has joined #opensolaris[21:00:07] <cslamar> hello, is it possible to expand a zfs block devices?[21:00:19] <tsoome> zvols? yes[21:00:31] <longcat> sure. and if you expand all of the block devices in a vdev, and set auto-expand on, it will grow too[21:00:39] <tsoome> zfs set volsize[21:00:59] *** Hedonista has joined #opensolaris[21:01:00] <cslamar> i mean if you create a virtual block device from a zpool can you expand that?[21:01:23] <tsoome> that was my answer:P[21:01:40] <cslamar> ok, sorry still kind of new with zfs[21:01:55] <glance> zfs create -V 2g tankpool/volume[21:02:08] <glance> zfs set volsize=3g tankpool/volume[21:02:44] <cslamar> ah[21:02:47] <cslamar> thanks :)[21:03:57] *** cslamar has left #opensolaris[21:04:24] *** twilling has quit IRC[21:04:24] *** vraa__ has quit IRC[21:08:49] *** yippi has quit IRC[21:09:42] *** yippi has joined #opensolaris[21:10:05] *** pgr has joined #opensolaris[21:13:20] *** icarus902 has quit IRC[21:16:42] *** forquare has quit IRC[21:16:46] *** syamajala has quit IRC[21:17:25] *** timeless_mbp has quit IRC[21:19:24] *** eleusis has quit IRC[21:19:36] *** jpd1234 has joined #opensolaris[21:21:00] *** eleusis has joined #opensolaris[21:22:35] *** Mdx4 has joined #opensolaris[21:25:20] *** nitrile has joined #opensolaris[21:27:46] *** twisti has quit IRC[21:30:15] <sickness> hi Mdx4 :)[21:30:29] <Mdx4> hi sickness :)[21:32:46] *** Calibellus has joined #opensolaris[21:33:26] *** Beket has quit IRC[21:35:14] *** nitrile has quit IRC[21:37:00] *** nitrile has joined #opensolaris[21:41:34] *** Hedonista has quit IRC[21:45:47] *** gdamore has quit IRC[21:46:25] *** pgr has quit IRC[21:49:39] *** ingenthr has quit IRC[21:50:00] *** ingenthr has joined #opensolaris[21:50:08] *** biant92 has left #opensolaris[21:50:32] *** Mace_N900 has quit IRC[21:54:05] *** TommyTheKid has joined #opensolaris[21:55:15] <TommyTheKid> is there some reason there is a network/dns/bind and a service/network/dns/bind ?[21:58:23] *** mbreitba has joined #opensolaris[21:58:45] <mbreitba> Hey guys have some questions about upgrading 2009.06 to the latest development build[21:59:01] <mbreitba> I'm not super savvy, and every time I try to upgrade, it kills off my destkop[21:59:18] <mbreitba> also, my desktop is stuck at 1280x1024[21:59:26] <mbreitba> any ideas on what could be causing my problems?[21:59:41] <TommyTheKid> there are a lot of "issues" noted in the release notes[22:00:24] <TommyTheKid> I'd recommend fresh installing with build 134 from http://genunix.org unless you have something you really really want to keep .. but my advice should be taken with a grain of salt, :)[22:00:52] <lewellyn> TommyTheKid: a fresh install won't get bugs filed :P[22:01:16] <lewellyn> mbreitba: if you have an xorg.conf, move it out of the way, on the new be.[22:01:31] <TommyTheKid> lewellyn: touche[22:01:49] <mbreitba> sweet - I didn't know that they had installers for the development builds[22:02:05] <lewellyn> mbreitba: the upgrade paths need to be tested at least as much ;)[22:02:08] <mbreitba> Will I still get a desktop environment with those?[22:02:14] *** snuff-home2 is now known as snuff-home[22:02:16] <lewellyn> yes[22:02:33] <TommyTheKid> I have just been fighting an issue with dns/server ... think I have found the "cause" .. but the cause is the /lib/svc/method/dns-server script :p[22:02:52] <mbreitba> excellent - time to blow out the old installation and start new[22:03:08] <lewellyn> TommyTheKid: file bugs :D[22:03:11] <lewellyn> mbreitba: why?[22:03:11] <TommyTheKid> or maybe something above that.. it seems like it wants to become the user "named" before ever starting the start method[22:03:14] <lewellyn> this isn't windows :P[22:03:29] <lewellyn> TommyTheKid: yeah. bet the method says to do that[22:03:36] <lewellyn> or manifest rather[22:03:59] <TommyTheKid> right, it does, I'd rather the method took a "run-as-user" config option instead of starting as the user :)[22:04:13] *** Seeder has joined #opensolaris[22:04:20] <lewellyn> did you look at the manifest yet? :)[22:04:23] <TommyTheKid> ... the reason being, if you run AS the user, it can't open /dev/poll RDRW, and fails exit 1 right away[22:04:52] <lewellyn> i would, but i'm finishing off an update to a support request before i make some phone calls *sigh*[22:05:00] <TommyTheKid> do you understand what I am trying to say (I seem to be still dizzy from being hungry earlier)[22:05:00] *** eff has joined #opensolaris[22:05:12] <TommyTheKid> sorry mbreitba I just sorta dove in over you :)[22:05:16] <lewellyn> TommyTheKid: i think you're saying the manifest is invoking the method as named[22:05:52] *** eff has quit IRC[22:05:55] <TommyTheKid> invoking the start method as the user specified in method_credentail[22:06:03] *** effnorwood has joined #opensolaris[22:06:16] <TommyTheKid> instead of having a propval name='run_as_user' ....[22:06:48] <lewellyn> um. that's about what method_credential does[22:06:53] <TommyTheKid> ... which would run the start method as "root" and then use the "-u USER" argument[22:06:58] *** effnorwood has quit IRC[22:07:33] *** nevdull has joined #opensolaris[22:07:48] <lewellyn> afaiui, it runs the script with the EUID matching the user specified[22:07:50] <TommyTheKid> I think the permissions on /dev/poll were wrong on my test[22:07:57] <lewellyn> well then[22:08:00] <TommyTheKid> maybe ignore this for a sec[22:08:15] <TommyTheKid> I was just trying to re-prove this problem on OS after bashing my head in fixing it on S10[22:08:40] *** effnorwood has joined #opensolaris[22:09:01] *** Amse has joined #opensolaris[22:09:03] *** corretico has quit IRC[22:09:28] *** effnorwood has quit IRC[22:09:43] *** effnorwood has joined #opensolaris[22:10:17] *** effnorwood has quit IRC[22:10:23] *** nevdull has quit IRC[22:10:48] <TommyTheKid> either way, it works when the process starts as root and then chowns itself, but didnt when it started as the user, but the real problem was that my configuration came from a far older version and I tried to quick-hack it for 9.3.6-P1 by quickly mknod'ing the poll device in the chroot, and failed to fix permissions[22:11:22] <TommyTheKid> so there is a difference between the two ways of starting as non-root.. but my stupidity caused it :)[22:14:30] <TommyTheKid> are all the "issues" upgrading from 2009.06 fixed for 134? I haven't tried admittedly :)[22:14:50] <lewellyn> hopefully they're fixed in 134a[22:15:06] <TommyTheKid> a? first I have heard of said build[22:15:19] <program> I tried to upgrade from 2009.06 yesterday and had some known issues with ICEauthority on Xorg startup[22:15:22] <program> but that was it[22:15:39] <program> decided to just do a fresh install of 134[22:16:20] <TommyTheKid> I do suppose we should test that upgrades work from at least the last couple full releases[22:16:22] <lewellyn> TommyTheKid: you think they were going to just use the dev build as-is for the release?[22:16:37] <lewellyn> if so, it would have been out long ago :P[22:16:39] <TommyTheKid> is 134a RC2 then?[22:17:00] <lewellyn> there have been no RCs yet[22:17:26] <TommyTheKid> lewellyn: I have not been involved in any of the tesing since 134-dev came out... things have been very quiet :)[22:17:29] <lewellyn> i've not actually seen confirmation there will be, either. but that'd be less-than-ideal, in my book.[22:17:36] <lewellyn> TommyTheKid: lurk moar[22:17:43] *** Phoenixland has quit IRC[22:18:03] *** rulereric has joined #opensolaris[22:19:03] *** radsy has joined #opensolaris[22:19:39] <TommyTheKid> I'd love to see RC's I think they raise interest and help clear up install/upgrade bugs... too many of us bounce from dev build to dev build and don't experience what a normal end user does ;)[22:19:58] *** piwi has joined #opensolaris[22:21:22] <lewellyn> i upgrade 111b->dev at least once or twice a week. that's why i grumble about 134 so much. it's somewhat broken for that upgrade[22:21:43] <TommyTheKid> not to say that any Open Solaris users are "normal end users" :)[22:23:10] <alanc> 134a packages are all built & delivered from the individual consolidations - they're doing the work now to combine them into iso's & repos and such[22:23:26] <TommyTheKid> sweet![22:24:22] <TommyTheKid> fresh 134 and sunray 4.3 are quite nice, that has been my "primary" desktop for a few weeks now[22:25:22] *** hsp has quit IRC[22:25:38] *** mroconnor has quit IRC[22:25:45] *** darrenb` has joined #opensolaris[22:26:14] <program> I wouldn't mind doing something like that, as my new house is cat5e wired[22:26:20] <program> how much can you pick up a sunray for?[22:27:29] <TommyTheKid> I would recommend getting a Sun Ray 1G minimum[22:28:00] *** darrenb has quit IRC[22:28:05] <TommyTheKid> I like the 2FS, I like to use a 24" monitor, and the regular Sun Ray 1 and Sun Ray 2 are limited to 1280x1024[22:28:19] <TommyTheKid> the SR3Plus jst came out, but I haven't seen one yet[22:28:50] <program> i'd like it as a little terminal in a mutual part of my house[22:28:58] <program> ya know, for guests and other misc uses[22:29:16] <TommyTheKid> once you get used to the silence, its hard to use a normal workstation again ;)[22:29:32] <stallion_work> alanc: no respins?[22:29:46] <stallion_work> or is this a preview?[22:30:00] <alanc> huh? 134a is a respin of 134[22:30:41] <TommyTheKid> i don't suppose we will know if there will be a "b" until "a" comes out :)[22:31:01] <alanc> whether any further respins are needed depends on whether WOS RE can build isos & repos that work for installing/upgrading, and whether any stopper bugs are found once testing begins[22:31:39] <alanc> and whether 134a meets all the other ship criteria set by release manangement[22:37:06] <stallion_work> alanc: sorry, I meant a further respin[22:37:10] <stallion_work> 134a is a given[22:37:35] <alanc> 134b has been discussed as a possibility, but at the moment is TBD[22:37:42] <stallion_work> gotcha[22:38:45] <lattera> I'm just excited for the release so that /dev can continue[22:38:57] <lattera> /dev is all I care about, heh[22:39:40] * lewellyn really hopes 134a is seeded as an RC[22:40:41] <tomww> if /usr/gnu/bin/cc is renamed to /usr/gnu/bin/gcc, then probably .-)[22:41:08] <lewellyn> has that bug been marked as fixed? :)[22:41:17] <tomww> is it a bug?[22:41:42] <lewellyn> well, it's a bug, but not necessarily one that's been filed[22:44:12] *** yippi has quit IRC[22:44:41] *** jthunder_ has quit IRC[22:45:47] *** dalibor has joined #opensolaris[22:49:32] <tomww> so what should be the bug description. link in gnu/bin/cc leads to mixing compilers gcc and SunStudio. or duplicate filename for compiler call.[22:50:45] <tomww> or simply, "it's been like thet never before".[22:51:07] <TommyTheKid> it should mix compilers, your path should point to one or the other?[22:51:18] <TommyTheKid> not that I agree with it being simple "cc" .. :)[22:51:23] <tomww> well, we need a strong reason why it is a bad idea to name ot simply "cc"[22:51:42] *** twisti has joined #opensolaris[22:52:00] <tomww> TommyTheKid: if GNU gcc and g++ and SunStudio cc and CC are on the same Path, regardless of the order of the two[22:52:23] *** thepickle has quit IRC[22:52:32] <tomww> directories, then everything is fine. We had that for many years and almos no mixing up of compilers occured.[22:53:05] *** scottdickson has joined #opensolaris[22:53:29] <TommyTheKid> right[22:53:37] <tomww> today on IPS based osol, if you put /usr/gnu/bin in the before the SunStudio your PATH, the wrong compiler is picked up if the script (e.g. configure) detects it runs on Solaris, then it thinks "cc" will give SunStudio[22:53:43] <lewellyn> tomww: argue that suncc is the superior compiler and with the default $PATH, it doesn't get used. and when linking C++ (and perl modules) that can be bad.[22:54:12] <lewellyn> good configure scripts check whether cc is gcc or not[22:54:16] *** Gman has joined #opensolaris[22:54:17] *** ChanServ sets mode: +o Gman[22:54:21] <tomww> we have only bad ones[22:54:32] <lewellyn> yes. but those should be patched ;)[22:54:33] *** melbogia has quit IRC[22:54:42] <tomww> and IPS osol should be the superios OS with less hazzle to compile F/OSS on it.[22:54:48] <lewellyn> yes[22:54:50] <CIA-21> John Ojemann <John.Ojemann at Oracle dot COM>: 6918206 Packets double counted on with "call now" rules, 6918859 pools should track bytes as well as packets for better usability, 6921174 ippool -ld crashes if nodes are inserted with ioctl and policy rules are not in place[22:54:51] <CIA-21> Peter Shoults <Peter.Shoults at Sun dot COM>: 6443649 softtoken should honor $HOME, avoid getpwuid, 6884140 softtoken touches $HOME too soon[22:54:55] <TommyTheKid> thats a good argument lewellyn .. but again, if I was going to compile something and wanted to use SunCC, I would put it first in my path[22:55:03] <lewellyn> having to explain why you can't mix compilers for C++ isn't superior in the least[22:55:09] *** datadigger has quit IRC[22:55:13] <lewellyn> TommyTheKid: and you'd know to do so how?[22:55:24] <alanc> unfortunately, "less hassle to compile FOSS" usually means "cc = gcc"[22:55:28] <TommyTheKid> of course I am no dev, andI feel like building software on Solaris is fantastically more difficult than it should be :)[22:55:30] <lewellyn> configure scripts tend to be smart enough to find both studio and gcc[22:55:56] <tomww> alanc: *lol*[22:56:04] <tomww> yes.[22:56:13] <lewellyn> using /usr/gnu/bin/gcc instead of /usr/gnu/bin/cc has not broken anything in my testing[22:56:34] <tomww> but why does it then break SFE on IPS osol?[22:56:36] <lewellyn> in other words, every configure script i've tried that expects to find gcc knows to look for it as "gcc" as well[22:56:38] *** timeless_mbp has joined #opensolaris[22:56:38] *** timeless_mbp has joined #opensolaris[22:56:43] <lewellyn> tomww: it does?[22:57:03] <tomww> lewellyn: got you. if you use my bootstrap I then recommend mv /usr/gnu/bin/cc /usr/gnu/bin/gcc[22:57:07] <lewellyn> yes[22:57:09] *** dajhorn has joined #opensolaris[22:57:18] <lewellyn> which gets broken by pkg fix :P[22:57:35] <tomww> yes. we are back at the beginning, to file a bug.[22:58:30] <tomww> well, I think we need a really clever idea what to write into that bug to get it accepted.[22:58:30] <lewellyn> in any case, if a configure script doesn't look for gcc as "gcc" as well as "cc", the sources probably have deeper breakage which would require patching anyhow.[22:58:36] <high-rez> Is the preview release build 134 fairly stable?[22:59:00] <lewellyn> tomww: fwiw, many linux distros don't even ship a "cc"[22:59:11] <tomww> lewellyn: my experience is, if is worse to have /usr/gnu/bin/cc then to have /usr/gnu/bin/gcc[22:59:14] <lewellyn> gcc as "gcc" has lots of precedent[22:59:14] <cypromis> they walkj one instead of shipping it ?[22:59:17] <lewellyn> tomww: it is[22:59:25] *** dajhorn has quit IRC[22:59:27] <perlmongo> When I compile things, I want to *know* which compiler I'm using, be it on AIX, Solaris, HP-UX or OSOL. If GCC is called "gcc", then I know. If gcc is installed with the name "cc" (on another path), there is guessing involved. I don't like guessing.[22:59:30] <lewellyn> cypromis: no. they only have "gcc"[22:59:36] *** ultrasparc-viii has quit IRC[22:59:38] <tomww> ah well, then this should be written in that bug. "Linux does not ship a "cc", mostly "gcc"[22:59:54] <lewellyn> tomww: some linuxes ship both. some ship "cc". some ship "gcc".[23:00:23] <tomww> we are on SOlaris, and we have a superios compiler, named SunStudio and calld as "cc" .-)[23:00:27] <lewellyn> yes[23:00:42] <lewellyn> there's a naming collision which really is unecessary[23:01:07] <tomww> that's a good point, "unnecessary" naming collision[23:01:23] <cypromis> unnecessay shipping of gcc :D[23:01:33] * tomww remembers a bug with the words: "excessive cleverness"[23:01:36] <InTheWings> pkg update-image process is supposely better than ever ?!?[23:01:53] <TommyTheKid> except its pkg image-update ?[23:02:12] <InTheWings> It takes so long, that I totally forgot its name ![23:02:14] *** datadigger has joined #opensolaris[23:02:41] <lewellyn> tomww: i'm filing the bug right now. :)[23:02:55] <InTheWings> And the gui fails silently , crashing[23:03:07] <tomww> perlmongo: thanks, noted.[23:03:52] <TommyTheKid> oh thats too creepy, I just re-booted a system that was running REALLY old SNV (102) and when the live CD got to the place where it started X11, I saw the old login screen momentarily on the ILOM remote console[23:04:02] <TommyTheKid> it must have still been in video memory or whatever[23:04:24] <tomww> perlmongo: "guessing" and perl and "want to know which one" ... if you compile software and have not fully control of the build system without patching it[23:04:26] *** jiraia has quit IRC[23:05:20] <lewellyn> tomww: are you in front of a "stock" opensolaris right now?[23:05:22] *** merzo has quit IRC[23:05:23] <tomww> lewellyn: thanks for that bug filing. you have native language capabilities, which I haven't[23:05:42] <lewellyn> tomww: i expect you to provide feedback on the bug ;)[23:05:45] <tomww> lewellyn: this is SXCE build 130.[23:05:48] <lewellyn> bah.[23:06:02] <perlmongo> tomww: exactly[23:06:06] <lewellyn> guess i'll provision a new VM for myself in a bit[23:06:13] <lewellyn> i want to verify behavior :P[23:06:25] *** sk0rd has quit IRC[23:06:26] * tomww admits that he has used a spare partition on a second computer to install 134/IPS[23:06:44] <tomww> you could also use the tes.topensolaris.org build farm[23:06:49] * lewellyn deploys JeOS images to hyper-v[23:06:56] <lewellyn> 10 minutes to deploy, at most :)[23:07:01] *** Animal-X has quit IRC[23:07:07] <lewellyn> my zone isn't fresh in the test farm[23:07:15] <tomww> or you could move back gcc to cc on a know good build machine of yours[23:07:19] <lewellyn> i want to make sure i'm testing the default packaging[23:07:30] <lewellyn> i think the man pages are messed up too[23:07:40] <lewellyn> worse than the naming[23:07:51] <tomww> who needs manpages[23:07:53] * tomww runs[23:08:29] *** hajma has quit IRC[23:08:30] <TommyTheKid> no one reads the documentation no matter what format its in :)[23:09:33] <lewellyn> it's just yet another compelling argument, if i'm right... ;)[23:10:58] <TommyTheKid> did someone answer my question about the difference between network/dns/bind and service/network/dns/bind ?[23:11:07] <lewellyn> not i[23:11:46] *** Seeder has quit IRC[23:12:54] <lewellyn> ok. guess i'll switch machines and provision a new vm...[23:13:25] *** astra-x has quit IRC[23:13:42] <tomww> TommyTheKid: looks a bit strange that "service/network/dns/bind" . a previous experiment or somthing else like a very early build, ages old?[23:14:03] <vraa> i have finally gotten samba to work with the force masks and stuff[23:14:04] *** nachox has quit IRC[23:14:09] <TommyTheKid> you cant install the shorter one cause it says there are two matches[23:14:23] <tomww> lewellyn: I had problems with the gnu/bin/cc beast, but don't remember the spec file(s!) which made problems.[23:14:23] <TommyTheKid> or maybe I can't weild the pkg tool correctly :)[23:14:36] <vraa> http://osol.pastebin.com/KDLx2bPm <-- if anyone wishes to compare smb.conf's[23:14:49] <tomww> I think every spec file secifying that: CC=cc and then expection it to be the pure SunStudio compiler[23:15:36] <alanc> looks like pkg:/network/dns/bind is the client side (dig, nslookup, libraries) and pkg:/service/network/dns/bind is the server side[23:15:47] <CosmicDJ> vraa: http://blogs.sun.com/timthomas/entry/making_samba_go_faster[23:15:57] <alanc> at least from a quick look at pkg contents for earch[23:16:00] <alanc> each even[23:16:34] <alanc> unfortunately, without the pkg:/ prefix, pkg contents network/dns/bind helpfully lists the combined contents of both, which is confusing[23:16:52] <TommyTheKid> alanc: that makes sense[23:16:55] <lewellyn> ok. provisioned. let's install these compilers[23:16:59] *** raichoo has quit IRC[23:17:04] <tomww> vraa: your netbios alias should probably be DNS resolvable, like "getent hosts VKCSVR-FILES" and all your clients should succeed with "nslookup VKCSVR-FILES". Eventually switch on on the client sied: use DNS/TCP for netbios name resolution"[23:17:52] *** robo has quit IRC[23:17:56] <high-rez> For the .usb downloads of opensolaris - can I just use dd to put the image onto a usb flash drive? E.g. dd if=osol-dev-blah.usb of=/dev/sdc (in linux) ?[23:18:05] <high-rez> Or is there some special prep that needs to be done ?[23:18:07] <alanc> smrt: explain liveusb[23:18:08] <smrt> To make a Live USB version of OpenSolaris, you need a .usb image (genunix.org) and a way to put it on a USB drive. Usually, one uses http://src.opensolaris.org/source/raw/livemedia/livemedia/usbcopy (Solaris), http://chonan-en.blog.pid0.org/2009/11/how-to-create-opensolaris-live-usb.html (Unix/Mac), or http://devzone.sites.pid0.org/OpenSolaris/opensolaris-liveusb-creator (Win)[23:18:50] *** Mace_N900 has joined #opensolaris[23:19:46] <high-rez> Hmm[23:20:14] <rkeene> N900![23:20:21] *** baza11215 has quit IRC[23:20:50] <high-rez> Nokia N900?[23:21:10] <rkeene> That's my assumption.[23:21:20] *** radsy has quit IRC[23:21:33] * high-rez is lost[23:21:35] <lewellyn> gah. p.o.o is slow.[23:21:41] <high-rez> but i wish i had an n900 ;)[23:21:47] <rkeene> I wish I had one too.[23:21:53] <longcat> Heh. I mirror IPS, but you could hardly say it's faster[23:21:53] <alanc> planet or pkg?[23:21:59] <TommyTheKid> oh, p.o.o is running on a N900, no wonder its slow :p[23:22:07] <lewellyn> anyone have good directions on how to set up an http proxy for p.o.o?[23:22:11] <lewellyn> alanc: pkg[23:22:29] <longcat> search the webcvs or whatever, there's a http conf there. or ask #pkg5[23:23:03] * lewellyn shudders at asking #pkg5 :P[23:23:16] <longcat> cant imagine why, they seem like good peoples[23:23:30] <lewellyn> i think i tend to ask most of the time when the good ones aren't there :P[23:23:36] <vraa> tomww: sorry i dont understand what you are saying, i have a win2k8 box running my dhcp and dns server.[23:23:54] <longcat> search the webrev or whatever for http.conf, there's an example reverse proxy setup there. if that's what you're looking for[23:23:57] *** Yu\2 has quit IRC[23:24:18] <lewellyn> longcat: i'm just trying to find out if i need to do anything less-than-standard to make sure i cache properly[23:24:24] *** eleusis has quit IRC[23:24:35] <longcat> ok.[23:24:36] * longcat folds[23:24:38] <vraa> CosmicDJ thanks for the link, i will go through this, but my samba hasn't been too slow, mostly my bottleneck is the disk IO I think, only 2 drive raid1z array i have[23:25:00] *** albakry has quit IRC[23:25:08] <longcat> should just mirror if you have 2 drives, you cant expand that raidz vdev[23:25:37] <tomww> vraa: that it as kind of proactive, in case your clients can't find the server and such[23:25:56] *** eleusis has joined #opensolaris[23:25:56] *** eleusis has joined #opensolaris[23:26:25] * tomww wanders off, good night. lewellyn pls drop me the bug id later thx:)[23:26:33] <lewellyn> tomww: i will. :)[23:26:37] *** Mace_N900 has quit IRC[23:26:43] <lewellyn> installing 15000 files :D[23:26:52] <lewellyn> well, 15000 actions[23:27:58] <tomww> looks like a powerfull software system[23:28:34] <lewellyn> pkg(5)[23:29:12] *** melbogia has joined #opensolaris[23:29:24] <lewellyn> well, my hunch with man pages was wrong[23:29:48] <TommyTheKid> i still contend that a package should be ONE file, that would make the pkg servers job a lot simpler, and it would make the pkg tool's job quite a bit simpler as well :)[23:29:58] <TommyTheKid> of course it would suck for bloated packages :)[23:30:08] *** scottdickson has quit IRC[23:30:16] <Aria> Yeah, I wish there was a packed transport option[23:30:40] <lewellyn> Aria: pipelining + compression isn't enough?[23:31:09] <Aria> Hard to hack with a web browser to deal with cases where stuff's screwed up. Single files are kinda nice that way.[23:32:20] *** jollyd has joined #opensolaris[23:32:31] *** jollyd has left #opensolaris[23:32:46] <lewellyn> the many small file approach is far better for checksumming and refetching bad bits though[23:33:37] *** chainey has joined #opensolaris[23:33:47] *** scottdickson has joined #opensolaris[23:35:28] *** rcarcamo has quit IRC[23:36:11] *** corretico has joined #opensolaris[23:36:53] *** Dganitch has joined #opensolaris[23:37:22] <Dganitch> looking for some help with CIFS in opensolaris[23:37:30] <vraa> Dganitch welcome to the club :D[23:37:44] <Dganitch> hello[23:37:46] <Dganitch> thnx[23:38:03] <Triskelios> smrt: explain cifs server[23:38:03] <smrt> OpenSolaris has a new CIFS (Windows file sharing in SMB) server in the kernel. Good resources for it include http://dlc.sun.com/osol/docs/content/SSMBAG/troubleshooting.html and http://wiki.genunix.org/wiki/index.php/CIFS_Service_Troubleshooting . The CIFS provides better performance and easier administration for CIFS than Samba but Samba provides other SMB services (See also: samba)[23:39:18] <Dganitch> will look them over,..[23:39:23] <Dganitch> thnx[23:39:30] <vraa> there are two ways you can share using cifs, inbuilt kernel cifs and samba seperately[23:39:56] <vraa> i would say cifs doesn't offer easier administration, samba does[23:40:55] <kmays> vraa: See..I wasn't crazy after all about that ZFS ALC patch for samba. ;)[23:41:23] <vraa> Tresk showed a link which said nfs v4 acl support IS still in samba 3.0.37[23:41:23] <lewellyn> i was just about to ask if windows users can modify samba's ZFS ACLs via explorer[23:41:35] <CosmicDJ> vraa: so creating a smb.conf is easier than zfs sharesmb=on? ;)[23:41:42] *** RoyK has left #opensolaris[23:41:53] <vraa> CosmicDJ yeah, plus you can use SWAT and smbtree and smbutil[23:42:08] <lewellyn> CosmicDJ: and apparently it's easier to join a domain than smbadm join -u user domain[23:42:28] <vraa> i still haven't figured out how to do the domain joining part lewellyn[23:42:38] <lewellyn> smrt: explain cifs active directory[23:42:38] <smrt> The kernel CIFS server is capable of joining an Active Directory domain. See http://blogs.sun.com/timthomas/entry/configuring_the_opensolaris_cifs_server for instructions. See also: cifs workgroup; cifs and windows 2008 domain[23:42:39] <vraa> i tried to setup the krb5 client, but i kept getting an internal error, i'll save that for another day[23:42:51] <vraa> yeah that's the blog i followed[23:42:55] <lewellyn> those directions kind of suck but it's good enough to start with[23:43:09] <lewellyn> i keep meaning to write it up here. i'll probably do that this week some time[23:45:39] <Dganitch> swould anyone recommend Sanba over the cifs kernel?[23:46:03] <vraa> yeah i would[23:46:47] <lewellyn> Dganitch: if i needed to traverse datasets[23:46:50] *** synegy34 has quit IRC[23:46:59] <lewellyn> or to provide non-cifs services to windows clients[23:47:05] <vraa> lewellyn http://www.aspdeveloper.net/tiki-index.php?page=SolarisCIFSPermissions - maybe that's what you are looking for[23:47:24] <CosmicDJ> or to provide printer support for non-ipp windows machines[23:47:44] <lewellyn> vraa: that's with samba?[23:47:54] <lewellyn> CosmicDJ: that's non-cifs services ;)[23:48:28] <vraa> no that's with acl's, sorry i thought you were using zfs acl's[23:48:39] <lewellyn> yes. the cifs server makes that very simple[23:48:43] <vraa> err, it could be with samba or cifs, since samba supports acl's[23:49:08] <lewellyn> when i tried samba last, it wasn't manageable from windows very well[23:49:15] <lewellyn> maybe it's changed in the past year or so[23:49:38] *** Odin- has quit IRC[23:52:03] *** radsy has joined #opensolaris[23:54:48] *** sponix has joined #opensolaris[23:55:01] <CIA-21> Anders Persson <Anders.Persson at Sun dot COM>: 6943440 race in solookup() can cause smod_refcnt to be bumped multiple times for a single sockparams[23:55:02] *** nikolam has quit IRC[23:55:43] *** sergiusens has joined #opensolaris[23:55:43] *** nikolam has joined #opensolaris[23:57:20] *** eliyahu has joined #opensolaris