October 28, 2006 
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 | 31
October 28, 2006 [00:00:31] <gdamore> well i think so. obviously /usr/ucb/ps can find them. [00:00:46] <gdamore> i didn't know about pargs. kind of a pain, because i have to lookup the PID first. [00:01:09] <alanc> cd /proc ; pargs * [00:01:22] <dwc-> hm, pargs is new [00:01:38] <dwc-> well "new" [00:01:51] <gdamore> still, it would be handy to have a ps argument like -w under ucb [00:02:12] <alanc> ARC case for it was 2001, so that should be around S9 [00:02:29] <dclarke> $ which pargs [00:02:31] <dclarke> no pargs in /usr/xpg4/bin /sbin /bin /usr/sbin /usr/bin /usr/openwin/bin /usr/dt/bin /usr/ccs/bin [00:02:32] <dwc-> well, new relative to when I started using unix and solaris [00:03:09] <dclarke> okay .. its new to me also [00:03:15] <dclarke> its not in Solaris 2.5.1 [00:03:32] <alanc> it's on my S9 build machine, but not the S8 one [00:03:37] <dclarke> its not in Solaris 8 [00:03:39] <dwc-> isn's 2.5.1 eol'd yet? [00:03:56] <dclarke> its on some long long term support option still I think [00:04:06] <alanc> 2.5.1 is beyond dead - it's fully decomposed [00:04:11] <dclarke> for a million bucks a year you can get support for anything [00:04:16] <dwc-> 'vintage' support? [00:04:31] <dclarke> well .. its running here [00:04:42] <dclarke> and fully patched [00:04:50] <dclarke> not WABI patched however [00:04:51] <dwc-> when was the last time it got patches? [00:04:58] <dclarke> let me check [00:05:10] <g4lt-U60> bah, WABI was the coolest thing about 2.5.1 [00:05:49] <dclarke> I have 102571-12 from Nov 14 2003 [00:06:05] *** eboutilier has quit IRC [00:07:19] * gdamore has 2.5.1 source (unpatched) [00:07:44] <gdamore> thank god though, that it is only used for cases of historical interest. [00:07:54] <dclarke> pretty much [00:08:06] <dclarke> but I can run Windows 3.1 in it [00:08:13] <dclarke> and that means .. solitaire ! [00:08:17] <dclarke> :-) [00:08:24] <gdamore> win3.1... wwwww.... [00:08:35] <dclarke> and really really old Lotus Notes [00:08:42] <dwc-> and minesweeper [00:09:02] <dclarke> but the real kicker .. the reason why WABI makes it all worthwhile [00:09:14] * gdamore remembers running desqview/X with win 3.1. :-) [00:09:20] <dclarke> is that when the screensaver kicks in .. it does the flying stars on the whole screen [00:10:47] <dclarke> want to see a real piece of computer history ? [00:11:17] <dclarke> the absolute coolest thing to still have running from Intel ? [00:11:35] <dclarke> SunOS Release 5.5.1 Version Generic_103641-42 [UNIX(R) System V Release 4.0] [00:11:35] <dclarke> Copyright (c) 1983-1996, Sun Microsystems, Inc. [00:11:35] <dclarke> mem = 196216K (0xbf9e000) [00:11:35] <dclarke> avail mem = 185208832 [00:11:35] <dclarke> FP hardware exhibits Pentium floating point divide problem [00:11:36] <dclarke> If you wish to disable the FPU, edit /etc/system and append: [00:11:38] <dclarke> set use_pentium_fpu_fdivbug = 0 [00:12:04] <dclarke> I have an actual broken FPU Intel Pentium P90 processor here [00:12:45] <gdamore> are you saving that for historic/nostalgic value? [00:12:50] <dclarke> yep [00:13:02] <alanc> or for extra heating during those cold canadian winters? [00:13:15] <dclarke> I always had this dream that one day I'd call up Intel and tell them that I want them to replace my broken CPU for free [00:13:21] <dclarke> which .. they promised they would [00:13:25] <andersmo> historic, hysteric... =) [00:14:04] <gdamore> i gotta think there was a time limit on that offer. [00:14:14] <dclarke> yep .. most likely [00:14:20] <dclarke> but this machine will not die [00:14:30] <dclarke> it cost me a real bundle .. I mean a ton of money [00:14:33] <gdamore> of course, they'd probably want to do the exchange and write it up in some press release, at this point. :-) [00:14:40] <dclarke> I think I have theinvoice still [00:14:56] <dclarke> all in .. about $18,000 [00:15:13] <gdamore> wow. that's real money. half a thumper. :-) [00:15:16] <dclarke> for a dual CPU P90 with onboard SCSI and PCI and IDE and ISA and ... [00:15:33] <gdamore> *dual* P90? I didn't know they even existed. [00:15:41] <dclarke> it did and it does [00:15:45] <g4lt-U60> there were dual 486es [00:15:55] <dclarke> heck .. remember Himalaya ? [00:16:04] <dclarke> 16 processor boxes [00:16:05] *** svoboda has quit IRC [00:16:15] <gdamore> well there was dynix/sequent too on 386, but that was "special" [00:16:17] <g4lt-U60> HP put out one IIRC. I think they even put out a dual 386 [00:16:40] <dclarke> Compaq had the best quad proc boxes by 1997 [00:16:48] <dclarke> this box was made in 1994 or so [00:17:07] <dclarke> that explains the $18K price tag for the latest generation Pentium at the time [00:17:15] <dclarke> and .. the 192MB of RAM [00:17:31] <dwc-> yea, dual P90... I had one of those [00:17:32] * gdamore is trying to decide whether its worth ~ $350 to build his own VIA M1000 NAS system if it only has 10/100 and PATA. [00:17:58] <gdamore> the newer EPIA boards are aorund 2X the cost. [00:19:13] <mrdeviant> gdamore, http://www.logicsupply.com/product_info.php/cPath/78_93/products_id/545 [00:19:22] *** Symmetria has quit IRC [00:19:31] <mrdeviant> it's the 1.5ghz c7 with pata, sata and (crappy) gig-e. and it's only $250 [00:19:45] <mrdeviant> i've got one on my desk. i'm using it to write a kcf padlock driver. [00:19:56] <gdamore> right. i can get the M1000 for ~$150. [00:20:30] <gdamore> i actually like the EN12000 better, even though its "only" 1.2GHz, because it can run fanless. [00:20:35] <mrdeviant> i think the extra $100 is def. worth it. [00:20:37] <alanc> yeah - we had a Sequent at Berkeley with 20 386 CPU's [00:20:52] <gdamore> it was always having problems, IIRC. :-) [00:20:52] <alanc> was cool running gmake -j20 on [00:20:55] *** lolownia has quit IRC [00:21:00] <quasi> gdamore: you can even get up to 1.5 in fanless epia [00:21:09] <dclarke> sequent .. those were awesome machines [00:21:20] <gdamore> i wasn't in the CS depatment, so I couldn't get an account on soda. but i had a friend who had a login there. [00:21:23] <dclarke> I recall that Lotus was trying to get OS/2 running on one [00:21:39] <mrdeviant> gdamore, what about the cn10000eg ? [00:22:03] <gdamore> i thought the cn1000eg doesn't have gigE? [00:22:04] <alanc> soda wasn't that hard to get an account on - just had to claim you were interested in CS - didn't have to be a major [00:22:20] <mrdeviant> err yea, you're right. nix that then. [00:22:32] <alanc> of course they replaced the 20 386's with a single K2 and it got much faster [00:22:36] <gdamore> didn't know that at the time. i was doing sysadmin for the business school back then (early 90s) [00:22:55] *** jonkelle has quit IRC [00:24:24] * gdamore wonders if the phylon boards are any good [00:25:11] <quasi> gdamore: isn't phylon the same as jetway - http://linitx.com/product_info.php?cPath=12_138&products_id=1083 [00:25:36] <gdamore> no idea. on logic's site it looks like even the 1.5GHz have only 10/100 lan though [00:25:37] <quasi> gdamore: I hear reasonably well about them [00:26:28] <quasi> gdamore: with jetway you can extend with giga lan boards - there's a 1 port and a 3 port card [00:26:33] *** bengtf__ has joined #opensolaris [00:26:36] <richlowe> stevel: *finally* [00:26:48] <stevel> richlowe: ? [00:26:49] <gdamore> quasi: looking now. looks interesting [00:27:10] <richlowe> stevel: hg project proposal [00:27:16] <stevel> ahh.. yay :) [00:27:29] <gdamore> quasi: where does the board plugin, on the PCI slot? [00:27:51] <quasi> gdamore: I don't think it is [00:28:16] <gdamore> they don't _look_ like PCI [00:28:29] *** nwf has quit IRC [00:28:33] <quasi> gdamore: http://www2.multithread.co.uk/images/j7f2_eden_fanless_large.jpg - to the left of the power plug [00:28:38] <gdamore> anyone used any of these? [00:28:49] *** nwf has joined #opensolaris [00:29:11] <quasi> a friend of mine is pretty happy with his [00:29:20] <gdamore> whats he run on it? [00:30:53] <quasi> linux [00:31:20] *** Kernel86 has joined #OpenSolaris [00:32:30] <gdamore> http://linitx.com/product_info.php?cPath=12_138&products_id=1177 [00:32:37] <gdamore> this looks very, very cool [00:34:12] <gdamore> and it exists in 1.2 fanless too [00:34:52] <quasi> yeah, it does - and I bet if you took off that tiny fan and put an 8cm low rpm on it the case, you'd have no trouble cooling it [00:34:54] <stevel> wow, that was 4 +1's in 20 minutes [00:34:59] <stevel> that's gotta be the quickest i've seen :) [00:35:02] <gdamore> heh. the fanless models are about 30BP more than the faster 1.5 models [00:35:32] <sahafeez> its a nice board. i ran openbsd on it for a firewall [00:35:50] <gdamore> i think i'm gonna buy one. [00:36:06] <quasi> sahafeez: have you tried booting solaris on it? ;) [00:36:17] <sahafeez> no [00:36:45] <gdamore> it _looks_ like it has an RS232 on it too, in the picture, but the specs don't say so [00:37:08] <sahafeez> i am much to religious and think solaris on anything but sparc is evil ;) [00:37:38] <gdamore> i think anything but solaris on x86 is evil. :-) [00:37:51] <gdamore> that's true for sparc as well. [00:38:22] <gdamore> and (here's the kicker), I want to port Solaris to MIPS to make it true for MIPS as well. :-) [00:38:47] <Error_404> you and your MIPS... [00:38:56] <sahafeez> ha, i do not think i can get solaris 10 on my ss5. [00:39:05] <gdamore> i'm trying to convince mgmt that we should build a NAS device using Solaris, NFS/ZFS and VIA cpus. [00:39:14] <Error_404> sahafeez: i promise you you cannot [00:39:23] <Error_404> sparc32 support was removed in solaris10 [00:39:41] <gdamore> if we add power management and a builtin battery pack (builtin UPS!) it would be pretty cool. 1TB portable NAS. [00:40:03] <Error_404> gdamore: use laptop drives, they drink less power [00:40:12] <gdamore> we might. but they run slow. [00:40:20] <sahafeez> i have this great ss5, 110mhz, quad fastE, 256mb ram, 2x9gb and a pcmcia sbus card. it is sitting there. i just cannot bare to trash it yet i do nothing with it [00:40:27] <Error_404> but you need twice as many spindles, so it'll run faster [00:40:30] *** svoboda has joined #opensolaris [00:40:37] <quasi> gdamore: you can get 7200 rpm laptop drives [00:40:38] <mrdeviant> there are 10k 2.5" sas disks. [00:40:45] <Error_404> sahafeez: solaris9'll run on it [00:41:02] <gdamore> well, if you want spend $$, there are CF devices that are fast, too. [00:41:12] <gdamore> (not just CF, but flash-based 32GB drives) [00:41:12] <sahafeez> bah, yes, and i can start it up and wait an hour for it to boot [00:41:13] <Error_404> that'd be pretty cool too [00:41:13] <kimc> looks like jetway have a few Add-on boards including 3-port gig-e http://linitx.com/product_info.php?cPath=12_145&products_id=1085&osCsid=4ef97baff97ebdac2497ba379ab48062 [00:41:21] <Error_404> see how small you can fit 1TB in [00:41:26] <kimc> looks like jetway have a few Add-on boards including 3-port gig-e http://linitx.com/product_info.php?cPath=12_145&products_id=1085&osCsid=4ef97baff97ebdac2497ba379ab48062 [00:41:30] <Error_404> gumsitx+CF drives [00:41:45] <gdamore> gumstix is ARM, IIRC [00:41:49] <kimc> opps [00:41:55] <dclarke> gdamore : fast compact flash memory devices ? [00:42:02] *** mkhl has quit IRC [00:42:10] <quasi> kimc: you're waay behind - 00:26 < quasi> gdamore: with jetway you can extend with giga lan boards - there's a 1 port and a 3 port card [00:42:11] <dclarke> dclarke : where ? Kingston memory ? [00:42:11] <gdamore> not technically compact flash, but NOR flash based devices, yes. [00:42:34] <kimc> eh sorry :) [00:42:38] <dclarke> is there a 4GB USB thumb drive now ? [00:42:38] <gdamore> crap, i need to look it up. I think it was Toshiba that was making them. [00:42:48] <dclarke> or 8GB I mean [00:42:59] <gdamore> you can get 8GB/16GB CF. [00:43:17] *** nwf has quit IRC [00:43:31] *** bengtf has quit IRC [00:43:47] <Error_404> slam 100 of them together, over a terabyte of storage in a 6 inch cube [00:43:49] <quasi> dclarke: yeah, 8G is not too difficult/expensive [00:44:18] <dclarke> I better go check out Kingston .. their stuff has a lifetime warrantee [00:44:18] <gdamore> E404: need to have good filesystem support. wouldn't want to run Linux on it. :-) [00:44:25] <Error_404> mind you, you could also stack 2 500gb drives on top of each other & get the same effect [00:44:31] <quasi> Error_404: which is fine if you're only reading - writing much to flash isn't much good [00:44:37] *** nwf has joined #opensolaris [00:44:48] <Error_404> probably be cheaper too [00:44:51] <gdamore> there are some new technologies that are supposed to be better. [00:44:54] <Error_404> s/probably/certainly [00:44:59] <gdamore> E404++ [00:45:06] <gdamore> (does that make it E405?) [00:45:22] <Error_404> only after you evaluate it [00:46:28] <gdamore> I like the idea of dual 1G, dual SATA, and using a PCI board to get additional external SATA expansion. [00:46:43] <gdamore> with dual 1G and 802.3ad/IP multipathing, this thing could really _smoke_ [00:47:10] <gdamore> do we have the crypto support for the via hardware crypto in solaris yet? [00:47:16] <mrdeviant> no [00:47:20] <mrdeviant> i am writing a driver for it now, though [00:47:25] <gdamore> if not, if someone gets me sample hardware, i volunteer to write the code. [00:47:34] <gdamore> mrdeviant beat me to it. darn. :-) [00:47:49] <mrdeviant> i'll post the code as soon as the bloody ip office here signs the SCA [00:48:00] *** evilstoy has quit IRC [00:48:02] <gdamore> who do you work for? [00:48:15] <mrdeviant> penn state university [00:48:55] <gdamore> hmm... what algos does the via support? can ZFS use any of it for hardware offload? [00:49:18] <mrdeviant> via does sha-1, and sha-256 on some boards. zfs has support for sha-256 [00:49:31] <mrdeviant> but currently the zfs code doesn't call into KCF for SHA-256. but that's an easy fix. [00:49:45] <gdamore> we should definitely fix it. [00:49:51] <mrdeviant> i imagine sun will fix it before the niagara 2 comes out, since n2 will have sha-256 on-die. if not, it's real easy to fix [00:50:23] <gdamore> i'd imagine so. i've not looked at the code. [00:50:24] <mrdeviant> http://cvs.opensolaris.org/source/xref/on/usr/src/uts/common/fs/zfs/sha256.c is the relevant file [00:50:52] <mrdeviant> the big problem with the via SHA implementation is that it wants the entire buffer in memory before you hash it [00:51:07] <mrdeviant> doing incremental hashing is a pain. there's a hack in openssl to make it work, but i'm not sure it would work in the kernel [00:51:37] <dclarke> that static const there " static const uint32_t SHA256_K[64] " looks curious [00:52:25] <gdamore> take a look at any real symmetric alg, and you'll find big constant tables like that [00:52:25] <dclarke> * This is a very compact implementation of SHA-256. 38 * It is designed to be simple and portable, not to be fast. [00:52:39] <dclarke> interesting ... [00:52:54] <mrdeviant> there's an open bug to fix it. i don't have the number off-hand. [00:53:02] <gdamore> I should talk to my buddy Ferenc. He might have some optimizations for SPARC v9 SHA256. [00:53:11] <dclarke> I would think that this would be hand rolled assembly and not just C code [00:53:32] <gdamore> (This is the same guy that got several factors of improvement in sparc hand assmbler optimization/scheduling for various crypto algos) [00:53:33] <dclarke> it should be possible to get th thing all in L1 cache or all in registers [00:53:59] <dclarke> fascinating [00:54:15] <jamesd> depends on the hardware.. i bet on x64 and usparc it will go quite easily in cache and registers [00:54:28] <gdamore> that implementation is incredibly compact. [00:54:48] <gdamore> much smaller than i'm used to for e.g. MD5 or SHA1 [00:55:14] <dclarke> back to USB memory devices .. I see that Kingston is still topped out at 2GB "DataTraveler II" devices [00:55:32] <mrdeviant> right. compare it to the KCF sha2 module - http://cvs.opensolaris.org/source/xref/on/usr/src/uts/common/crypto/io/sha2_mod.c [00:55:55] <mrdeviant> granted, that file does sha-{256, 284, 512}, as well as the hmac of those [00:56:25] <gdamore> hmac isn't much, but the different key sizes and hmac variants add a bit [00:56:48] <gdamore> there's also the uio and mblk variants [00:57:09] <mrdeviant> yup. it's just lots of extra code to handle minor variations [00:57:12] <mrdeviant> but it adds up [00:58:04] <gdamore> the compactness of the zfs implementation probably helps a lot -- keep the cache warm [00:58:25] *** laca has joined #opensolaris [00:58:49] <gdamore> although, its not as compact as you first think if you expand those macros.... :-) [01:00:12] *** coffman has quit IRC [01:02:24] *** jmcp has quit IRC [01:03:38] *** gm152 has joined #opensolaris [01:03:57] *** cmihai has quit IRC [01:04:12] *** gm152 has quit IRC [01:04:26] *** gm152 has joined #opensolaris [01:04:32] *** gm152 has quit IRC [01:05:15] *** jmcp has joined #opensolaris [01:06:33] *** lplatypus has joined #opensolaris [01:07:26] *** jsubl2 has joined #opensolaris [01:09:10] *** bengtf__ has quit IRC [01:10:43] *** alobbs has quit IRC [01:12:13] *** nwf has quit IRC [01:14:04] *** lplatypus has quit IRC [01:14:59] *** Fish-- has joined #opensolaris [01:16:49] *** polk__ has quit IRC [01:19:15] *** mrdeviant has quit IRC [01:21:18] *** nwf has joined #opensolaris [01:23:48] *** Fish- has quit IRC [01:24:33] *** [1]Cyl has joined #opensolaris [01:28:01] *** Fish-- has quit IRC [01:30:20] *** sahafeez has quit IRC [01:31:05] *** piwi has joined #opensolaris [01:32:03] *** jengelh has quit IRC [01:35:07] *** galt has joined #opensolaris [01:35:12] *** g4lt-mordant has joined #opensolaris [01:35:20] *** g4lt-mordant has quit IRC [01:35:59] *** Mazon is now known as mazon [01:43:59] *** Cyl has quit IRC [01:43:59] *** [1]Cyl is now known as Cyl [02:00:46] *** Symmetria has joined #opensolaris [02:00:47] <Symmetria> hey all [02:00:49] <Symmetria> anyone awake? [02:01:16] <gdamore> yes. [02:01:30] <delewis> if you count reading Oracle documentation as being awake, I suppose I am, as well. [02:01:35] <richlowe> mostly [02:01:36] <Symmetria> gdamore know much about zfs? specifically, after reinstalling a box, how do I remount a zfs file system I have on another disk [02:01:47] <Error_404> zfs import [02:01:49] <delewis> Symmetria: zfs import [02:01:59] <delewis> specifically, zfs import <name of the pool> [02:02:07] <richlowe> zfs import on it's own will list what it can find. [02:02:20] <gdamore> ooohh oooh,.. .i know... ."zfs import"! :-) [02:02:38] <richlowe> haha [02:02:39] <richlowe> wow. [02:02:41] <richlowe> zpool import :) [02:03:10] <Symmetria> hrm [02:03:11] <Symmetria> config: [02:03:11] <Symmetria> apool ONLINE [02:03:11] <Symmetria> c2t2d0 ONLINE [02:03:11] <Symmetria> ok [02:03:21] <Symmetria> now how do I mount that automatically etc [02:03:34] <delewis> looks like you've already got it "imported" as such, just not mounted [02:03:49] <richlowe> zpool import apool should do the trick [02:04:04] <delewis> will it show the disk as online if it isn't imported? [02:04:11] <Symmetria> cannot import 'apool': pool may be in use from other system [02:04:21] <richlowe> delewis: not sure, I don't have an exported pool, nor one I *can* export. [02:04:22] <Symmetria> something claims its using it so I presume it is imported, just not mounted [02:04:26] <delewis> Symmetria: try using zfs mount [02:04:46] <delewis> richlowe: same here -- the only ZFS pools I'm near are my fileserver's pools and my workstation (which I am now) with my home directory ;-) [02:04:47] <richlowe> delewis: but yeah, I believe it shows ONLINE as long as the device is there. [02:06:24] <Symmetria> apool 3.8T 2.6T 1.2T 68% /mirror/iscsi-san-1 [02:06:27] <Symmetria> bingo [02:06:36] <Symmetria> thanks guys :) [02:06:42] <delewis> sure :-) [02:07:04] <Symmetria> heh now I just gotta get the damn system to see my other 2 san's properly [02:07:07] <Symmetria> and Im all done [02:07:22] * richlowe tries to tie a simple device name, and 3.8T together somehow. [02:07:40] <richlowe> I guess I'm used to WWN's in that context. [02:09:35] <Symmetria> heh richlowe 3.8 TB is because its an iSCSI SAN [02:09:44] <Symmetria> iSCSI san shows up as a single disk [02:09:54] <Symmetria> when in actual fact its just a large raid array [02:11:55] <richlowe> Yeah, I'm just used to FC-like device naming in that case :) [02:12:09] <richlowe> (or is it mpxio that gives you disk names from hell?) [02:12:30] <twincest> symm: you're still working on this? [02:12:38] <twincest> i thought you'd fixed it before [02:12:39] <Symmetria> twincest I finally actually migrated the main machine [02:12:46] <Symmetria> and hit *MAJOR* shit when I did [02:13:01] <Symmetria> solaris hated that hardware almost more than linux and fbsd did [02:13:18] <Symmetria> it flipped out about the lsi card etc [02:13:32] <twincest> maybe we can conclude that the hardware isn't very good :) [02:14:16] <Symmetria> twincest nah, it was just lack of drivers in this case [02:14:45] <twincest> rich: my c3t0d0s0 is 1T out of a 4T device, large devices are not restricted to fc :) [02:15:07] <richlowe> c5t600C0FF000000000098FD5275268D600d0 [02:15:08] *** lplatypus has joined #opensolaris [02:15:09] <richlowe> that kinda deal. :) [02:15:40] <richlowe> and yeah, it's mpxio, and I totally need caffeine. :) [02:16:00] <Symmetria> twincest see msg :) [02:16:27] <twincest> let me see if i can find which window you're in [02:16:47] <richlowe> twincest: I did say I guess I'm *used* to seeing WWN's, not that there weren't alternatives :) [02:16:51] <richlowe> it just looks strange without them :) [02:17:05] <richlowe> (WWN or similar, since I guess that may not be...) [02:19:30] *** vipe has quit IRC [02:20:40] *** gisburn has joined #opensolaris [02:25:43] *** vipe has joined #opensolaris [02:26:48] * gisburn watches this channel being boring again. [02:28:25] <gisburn> Ok, little quiz: [02:28:27] <gisburn> "I'm struck and cut, shaped and cooled, then bound by rings to release whats stored." - what am I ? [02:28:51] <twincest> ksh93 [02:29:27] <gisburn> umpf [02:29:30] <gisburn> funny [02:29:33] <gisburn> twincest: no. [02:29:38] <twincest> :( [02:29:39] <gisburn> twincest: try again, please [02:31:41] <twincest> ale barrel [02:32:07] <dclarke> a diamond [02:32:25] <gisburn> no [02:32:43] <dclarke> 4140 high carbon steel [02:33:02] <gisburn> dclarke: ... release what stored ? [02:33:13] <gisburn> dclarke: it may be made of steel, right... MAY.. [02:33:18] <dclarke> I am allowed a wild ass guess [02:33:36] <gisburn> dclarke: granted. ass the guess. [02:33:51] <gisburn> s/ass/ask/ [02:33:55] <delewis> http://www.geekpatrol.ca/blog/161/ [02:33:56] <gisburn> er [02:33:57] <delewis> sweet. :-) [02:34:28] * delewis takes back his statement about gcc (even generally) being faster on AMD64. [02:34:29] <gisburn> come on... three letters! [02:34:39] <delewis> than Sun Studio, that is. [02:34:48] <dwc-> ksh, without the 93? [02:34:54] <gisburn> dwc-: no. [02:35:08] <delewis> this benchmark pretty much says otherwise, though, it might be more applicable if the user benchmarked gcc *and* Sun Studio performance on each of the respective operating systems. [02:35:11] <gisburn> dwc-: it's a very very old riddle. [02:35:26] <dwc-> I guess I'm not old enough [02:35:45] <gisburn> dwc-: I am not that old either. [02:36:01] <gisburn> Ok, hint: SG-1 recently abused the riddle in the show. [02:37:11] <dclarke> what is SG-1 ? [02:37:18] <delewis> StarGate [02:37:19] <gisburn> dclarke: StarGate [02:37:33] <dclarke> I must be too old [02:37:40] <dwc-> well, google tells me the answer [02:37:44] <richlowe> it doesn't require a hint, just google the exact phrase. [02:37:45] <dwc-> but that's not very interesting now, is it [02:37:53] <dclarke> its cheating [02:37:54] <richlowe> dwc-: who needs interesting? :) [02:37:57] <dwc-> nor do I feel like yelling out the answer [02:37:58] <dclarke> and I am not doing that [02:38:00] <dwc-> hwen I googled it [02:38:13] <richlowe> indeed. [02:38:15] <dwc-> dclarke: exactly. [02:38:35] <dclarke> damn . .struck then cut [02:38:41] <dclarke> that makes me think a gemstone [02:38:47] <dclarke> cooled bug me [02:38:48] *** laca has quit IRC [02:39:16] <dclarke> then .. bound to rings [02:39:23] <dclarke> to release what's stored ... [02:39:33] <dwc-> the last part threw me... [02:39:34] <dclarke> is this an object or a concept ? [02:39:37] <richlowe> "by rings" not "to rings" [02:39:38] <dwc-> object [02:39:46] <gisburn> hint: everyone has it. it's needed for daily life. [02:39:56] <dclarke> oh damn ... [02:40:02] <dclarke> everyone has it ... [02:40:11] <dclarke> this is really an excellent riddle [02:40:24] <gisburn> You're using it at least twice a day, but usually much more often. [02:40:28] <dclarke> can you state it again please ? [02:40:38] <dwc-> dclarke: don't you have scrollback? [02:40:45] * dclarke is lazy [02:40:48] <gisburn> Ok, little quiz: [02:40:50] <gisburn> "I'm struck and cut, shaped and cooled, then bound by rings to release whats stored." - what am I ? [02:41:14] <stevel> 3 letters? [02:41:22] <gisburn> yes [02:41:23] <stevel> i suppose if you're into kinky stuff it could be your ASS [02:41:32] <gisburn> no. [02:41:37] <dwc-> hm [02:41:38] <stevel> well to each his own :-P [02:41:45] <dclarke> geez ... [02:41:50] * dwc- relays this interesting tidbit to esther [02:41:52] <dclarke> twice a day .. huh ? [02:41:53] <gisburn> stevel: even the "ksh" guss was much closer than "ass" [02:41:54] <richlowe> stevel: did ass really need to be uppercase? :) [02:42:03] <stevel> richlowe: very much so :) [02:42:08] <richlowe> I find 'ksh' and 'ass' are pretty close. [02:42:09] <gisburn> dclarke: well, you leave and come back. [02:42:09] <dwc-> ASS is very different than ass [02:42:18] <dclarke> what ? [02:42:19] <gisburn> dwc-: ?! [02:42:26] <dclarke> leave and come back ... [02:42:40] <dclarke> this is killin' me [02:43:00] <dclarke> at least twice a dya .. not zero ? [02:43:09] <dclarke> twice a day ... [02:43:12] <gisburn> dclarke: only when you do not leave once a day. [02:43:15] <dwc-> well, it depends on what you do all day [02:43:36] <gisburn> dclarke: the rings can hold multiple XXX. [02:43:41] <dclarke> plunk my ass, lowercase ass, in a chair in front of a damn computer [02:44:35] <dclarke> struck then cut ... [02:44:43] *** Gadzooks has joined #opensolaris [02:44:49] <stevel> ohhhhhh [02:44:56] <gisburn> dclarke: it describes the old way how these items were made. [02:44:57] <delewis> you don't strike a gemstone [02:45:03] <dclarke> I'm struck and cut, shaped and cooled, then bound by rings to release whats stored. [02:45:04] <delewis> struck and cut sounds like making coinage [02:45:06] <dwc-> it's my hot body. struck and cut. [02:45:15] <Gadzooks> gnome-pilot is needed by someone. [02:45:19] <delewis> as does being bound by rings [02:45:20] <Gadzooks> dclake <-- [02:45:30] <stevel> dwc-: remind me next time i see you to whack you with a baseball bat and then stab you with a knife ;-) [02:45:42] <dclarke> hello there [02:45:51] <dwc-> =) [02:45:57] <dclarke> Gadzooks: thor and ra are fully up to date [02:46:02] <Gadzooks> hi..smokes if ya got em. [02:46:21] <Gadzooks> yes..but gnome-pilot is missing. [02:46:30] <dclarke> Gadzooks: one sec [02:46:33] <dwc-> hm, so delewis is onto metallurgy [02:46:38] <richlowe> stevel: I'm struggling to understand sch's bit about bundles. [02:46:44] <dclarke> Gadzooks: also .. osiris and isis have Studio 11 now [02:46:53] <dclarke> Gadzooks: need to be patched still [02:46:53] <richlowe> stevel: There's nothing you could break that is recoverable from a bundle, but not a local clone. [02:47:03] <Gadzooks> cool [02:47:04] <richlowe> (though the latter may take some use of strip) [02:47:05] <stevel> richlowe: i think he doesn't trust our stellar SCM environment ;-) [02:47:18] <richlowe> stevel: sod that. [02:47:24] <richlowe> stevel: I don't trust any of it, not just the SCM ;) [02:47:34] <stevel> i interpreted it to mean "if onnv-gate goes down for some reason, at least people can always download a bundle" [02:47:45] <gisburn> come on. [02:47:52] <gisburn> it's easy. [02:47:55] <gisburn> three letters. [02:47:58] *** rokerr has joined #opensolaris [02:48:06] <dclarke> ERROR: no matching SysV PKG found. [02:48:10] <stevel> richlowe: likewise, i don't trust any of it - but i actually care about the SCM stuff - so i'll bust my ass to get that working again versus something... well... less important :-P [02:48:28] <dclarke> gnome_pilot - The GNOME PalmPilot integration suite [02:48:28] <Gadzooks> ? [02:48:38] <dclarke> one sec [02:49:04] <movement> hrm 54 seconds to pull onnv-gate with the past week's or so changes [02:49:05] <movement> not bad [02:49:12] <gisburn> hint: 6B6579 [02:49:17] <dclarke> gnome_pilot - The GNOME PalmPilot integration suite [02:49:17] <dclarke> (sparc) 2.0.14 [02:49:23] <dclarke> going in now .. [02:49:29] <dwc-> what kind of obvious hint is that now.... [02:49:29] <dclarke> Installation of <CSWgnomepilot> was successful. [02:49:40] <dwc-> also 4B4B59 [02:49:42] <dclarke> Installation of <CSWgnomepilot> was successful. [02:49:48] <dclarke> done .. thor and ra [02:49:59] <Gadzooks> ok..will check..thanks [02:50:01] <dclarke> now I have to file a bloody change log report [02:50:08] * stevel has got the new opengrok server auto-updating from the companion, jds, and onnv-gate repositories every half hour now [02:50:09] <Gadzooks> ;> [02:50:12] <dclarke> Gadzooks: we done for now ? [02:50:24] <gisburn> dwc-: KKY ? [02:50:26] <movement> stevel: do we have a stable SWAN path for the hg repo? [02:50:27] <Gadzooks> yep [02:50:27] <richlowe> stevel: it wasn't meant as slight. [02:50:28] <stevel> once i figure out the best way to also restore g11n, sfw, and nws - i can put it into production [02:50:39] <richlowe> movement: I would hope not, no. [02:50:40] <dwc-> gisburn: can't type [02:50:41] <stevel> richlowe: i didn't interpret it as such :) [02:50:45] <richlowe> movement: we're all meant to use the same one. [02:50:50] <dclarke> Gadzooks: I'm struck and cut, shaped and cooled, then bound by rings to release whats stored." - what am I ? [02:50:54] <dwc-> or can't read [02:50:56] <dwc-> or maybe can't do math [02:50:57] <richlowe> stevel: the whole point of testing it is not trusting, but *checking* :) [02:50:58] <dwc-> one of the two [02:51:03] <richlowe> trusting it is useless :) [02:51:05] <dwc-> err three.... (must be the math) [02:51:08] <movement> richlowe: you don't have to deal with SWAN's proxies :/ [02:51:18] <richlowe> movement: Quite true. [02:51:21] <richlowe> movement: I have this kinda working, btw. [02:51:26] <stevel> movement: yeah... but only as a side effect of the mirror/bridge [02:51:37] <stevel> /net/tonic-gate.sfbay/builds/onnv_external [02:51:38] <richlowe> movement: not perfect, in fact, behaving in a way I like, but can't explain. [02:51:41] <richlowe> movement: but "different" [02:51:49] <movement> richlowe: oh? [02:51:52] <movement> stevel: ta [02:52:00] <richlowe> movement: lemme rebuild with the workaround gone, and I'll paste. [02:52:37] <stevel> it's stable in that that's what the bridge relies on. it may move somewhere better (onnv.eng would be a better place to house it honestly) in the future [02:52:53] <stevel> but it's fine where it is for now i suppose [02:53:07] <gisburn> noone figured it out ? [02:53:21] <stevel> gisburn: i did [02:53:44] <gisburn> my god, a channel full of experts and the brightest people at sun microsystems... and noone got it... ;-( [02:53:47] <Symmetria> enter selection: 5 [02:53:47] <Symmetria> Resetting HBA. This may take a few minutes. [02:53:48] * Symmetria snores [02:54:34] <Symmetria> heh iSCSI is super picky about its auth [02:55:27] <twincest> hmm, i reported a bug to intel 11 days ago and no answer yet [02:55:32] <twincest> i suppose it really is free support :) [02:56:01] <dwc-> support free? [02:57:43] <gisburn> hint: you need one for each of { house, office, car, bicycle, garage, etc } [02:57:54] <gisburn> (usually) [02:58:19] <stevel> bicycle???? [02:58:35] <gisburn> stevel: at least in germany if you want to keep it. [02:59:31] <dclarke> be right back [03:01:05] <spawrq> hmm... [03:01:14] <dwc-> I suspect you needed one at ucsd too [03:01:19] <spawrq> this solaris internals book doesn't seem to have any info on zfs in it. [03:01:22] <dwc-> I know anything unattached at berkeley would wander off [03:01:37] <richlowe> stevel: if you were going to provide an NFS accessible gate, I think I'd do it backward to that. [03:01:48] <stevel> richlowe: backward? [03:01:51] <richlowe> stevel: leave that as an implementation detail, and pull yourself a local clone. That way the changes fully go through os.o [03:02:02] <stevel> ah [03:02:03] <stevel> got it [03:02:42] <richlowe> it shouldn't make a difference, obviously, but any difference it does make would be good to catch, rather than hide :) [03:02:49] <stevel> yeah - we'll certainly keep a mirror once ON switches to the outside gate - i suppose we might as well start that mirror now [03:04:09] <dwc-> nfs accessible gate ... externally? [03:04:16] <stevel> alright i'm starving. i'm heading out for dinner [03:04:21] <stevel> adios and have a great weekend everyone [03:04:24] * dwc- waves [03:04:32] *** stevel has quit IRC [03:06:42] <jteo> morn * [03:06:51] <hile_> hey jeremy [03:06:58] <jteo> hello hile_. [03:09:26] <sickness> i'm back [03:10:16] *** lisppaste3 has quit IRC [03:13:13] *** lisppaste3 has joined #opensolaris [03:15:29] *** jengelh has joined #opensolaris [03:16:25] <Gadzooks> dclarke <- A ham ? ;> [03:16:47] *** rokerr has quit IRC [03:19:22] <Symmetria> heh [03:19:25] <Symmetria> with a bit of luck [03:19:33] <Symmetria> devfsadm should now find my other 2 disks [03:19:48] <Symmetria> then I'll have managed to get 38 outta 42 disks back online [03:21:18] <alanc> gisburn: the brightest people at Sun Microsystems are smart enough not to be here after 6pm on a Friday night - you're left with the rest of us now [03:21:58] <Symmetria> hrm how long does devfsadm -v normally take [03:22:16] <delewis> Symmetria: depends [03:22:39] <delewis> mostly on what sort of devices and how many of them you have attached to the system [03:22:52] <Symmetria> but I presume 15 minutes = something wrong? [03:22:54] <delewis> in the case of SAN-attached storage, it could take awhile [03:23:00] <icon> evening guys [03:23:06] <delewis> hmm, that seems a bit lengthy [03:23:14] <icon> finally getting real work started on the ports collection [03:23:17] <delewis> something could be wrong between the iSCSI initiator and target [03:23:18] * icon sighs contentedly [03:23:29] <Symmetria> heh I rebooted it :) same thing I did last time after getting the hba right for number 1 [03:23:30] <delewis> or it could just be slow if you're on a network with limited throughput. [03:23:34] <Symmetria> will probably come up fine [03:23:52] <delewis> I remember when I played around with iSCSI on my 100Mbps network, it took *forever* to detect the targets [03:24:25] <Symmetria> heh odd that it takes so long since the targets arent software and arent on limited bandwidth [03:24:28] <Symmetria> and the hba's already see them [03:24:45] <Symmetria> heh will see what happens on reboot [03:25:55] <Symmetria> heh would still love to know why this version of solaris point blank refuses to boot off my vga card [03:26:00] <Symmetria> its console boot or nothing [03:27:12] <icon> Symmetria: what device are you using? [03:27:41] <Symmetria> icon with the svcadm stuff? or in regards to the vga card? [03:27:42] <gisburn> what is the easiest way to print only the user of a file ? [03:28:12] <icon> vga card [03:28:28] <Symmetria> onboard thing on this intel woodbridge server board [03:28:30] <delewis> gisburn: you mean the owner? [03:28:47] <gisburn> delewis: yes [03:29:24] <delewis> ls -l | grep website | awk '{ print $3 }' [03:29:33] <delewis> er [03:29:37] <delewis> s/website/<file name>/ [03:29:48] <delewis> that's the most portable way, anyway :-) [03:29:49] <icon> Symmetria: thats a gma9xx isnt it? [03:30:45] *** gnu2it2 has joined #opensolaris [03:31:03] *** piwi has quit IRC [03:31:40] <delewis> (IIRC -- too lazy to look at the spec at the moment), POSIX defines the output of ls -l, so the owner of the file should always be the third-column. [03:32:40] <Symmetria> whats the zpool command to change a mount point of a pool again? [03:32:53] <Symmetria> icon Im not actually sure, not around the machine at the moment to check [03:33:02] <delewis> Symmetria: you can use zfs set [03:33:13] <delewis> specifically zfs set mountpoint=<mount point> <pool name> [03:33:53] <Symmetria> apool 3.8T 2.6T 1.2T 68% /mirror/iscsi-san-1 [03:33:54] <Symmetria> bpool 3.8T 24K 3.8T 1% /mirror/iscsi-san-2 [03:33:54] <Symmetria> cpool 3.8T 24K 3.8T 1% /mirror/iscsi-san-3 [03:33:57] <Symmetria> ! BANG! at last [03:34:08] * Symmetria does happy dance [03:34:16] <Symmetria> now its just the internal array thats not there yet [03:34:19] *** alanc is now known as alanc-away [03:34:51] *** packetjunkie has joined #opensolaris [03:36:22] <sickness> can sunray software 3.1 be installed on opensolaris? (like on sxcr b50 ?) [03:36:36] <Symmetria> hrm after installing a kernel driver package, do I have to reboot to load the driver or is there a command I can use [03:36:41] <delewis> sickness: if you're running Solaris Express, you shouldn't have a problem [03:37:01] <delewis> though, if you try installing it on any of the other OpenSolaris-based distributions you may encounter some difficulties [03:37:05] <delewis> especially with missing pkgs, etc. [03:37:57] <delewis> Symmetria: devfsadm should do that, I would think [03:38:25] <delewis> you could touch /reconfigure or append "-r" to your kernel boot arguments, but all that does is call devfsadm on boot [03:39:48] <BR_> Hey all, I have a ZFS question, we are currently in the progress of migrating some data from UFS to ZFS and I am noticing the utilisation of the caches on my disk arrays has gone right down. can anyone explain to me why this is? [03:40:07] <Symmetria> nah delewis I meant without a reboot to actually load the driver itself, but its cool, I bounced it anyway just to make sure everything else comes back right before bed [03:40:13] <delewis> BR_: ZFS disables the on-disk cache, IIRC. [03:40:20] <delewis> for data integrity reasons [03:40:41] <delewis> (most disks return successful write status once the data is written to cache, but not actually to the disk, itself) [03:41:06] <Symmetria> heh, I am such a dead man, gf is fast asleep and when I get up to go to bed she's gonna wake up and realize its 3:40am and I've been working all night again [03:41:13] * Symmetria thinks he is gonna get his ass kicked [03:41:23] <BR_> delewis: http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6322205 [03:41:29] <BR_> delewis: still think so? [03:42:11] <delewis> hmm, not sure [03:42:23] <BR_> delewis: :) Now you are in the same boat as me, enjoy the ride ;) [03:42:25] <delewis> you may consider searching through the zfs-discuss archives or sending a message to the list [03:42:26] * Symmetria waits while server runs through a gazillion checks before booting [03:42:30] <twincest> svn_35 is after the S10 integration [03:42:42] <delewis> that's true [03:42:43] <twincest> it might not have made it back [03:42:54] <delewis> in fact, I doubt it did [03:43:00] <BR_> hrmmm... [03:43:07] <delewis> S10u3 will probably have it, though [03:43:14] <delewis> you'll definitely want to confirm that [03:44:40] <sickness> delewis: tnx! should sunray server 4 work too? [03:44:59] <delewis> sickness: I don't see any reason why not, but I'm just talking generally [03:45:08] <BR_> okies [03:45:10] <BR_> http://sunsolve.sun.com/search/document.do?assetkey=urn:cds:docid:1-1-6322205-1 [03:45:12] <delewis> unless there's a bug, applications that run on Solaris 10 should run on Nevada. [03:45:15] <twincest> i thought srss has problems on SX [03:45:24] <BR_> http://sunsolve.sun.com/search/document.do?assetkey=1-21-122650-02-1 [03:45:25] <twincest> (no idea of the details though) [03:45:47] <sickness> oh, I see that 4 is nothing but a collection... ok tnx :) [03:45:48] <BR_> don't know if you can see them [03:45:53] <BR_> imagine you can [03:46:15] <twincest> br: if you have a contract can't you just ask sun? [03:46:27] <BR_> twincest: yes I can, but I thought I would ask here first :) [03:46:43] <BR_> twincest: but I will bother Sun on monday morning unless someone has a good answer :) [03:47:28] <Symmetria> dpool 1.1T 24K 1.1T 1% /mirror/internal [03:47:30] <Symmetria> YES!!!!!!!!!!!!!!!1 [03:47:33] * Symmetria does happy dance [03:47:38] <Symmetria> LSI internal raid back online [03:47:39] <jteo> gmm? [03:47:56] <Symmetria> all 13 active terabytes are back :) [03:48:13] <jteo> Symmetria: :) [03:48:26] <Symmetria> apool 3.8T 2.6T 1.2T 68% /mirror/iscsi-san-1 [03:48:26] <Symmetria> bpool 3.8T 24K 3.8T 1% /mirror/iscsi-san-2 [03:48:26] <Symmetria> cpool 3.8T 24K 3.8T 1% /mirror/iscsi-san-3 [03:48:26] <Symmetria> dpool 1.1T 24K 1.1T 1% /mirror/internal [03:48:32] <Symmetria> :))))))))))))))))) [03:48:51] <BR_> twincest: any bright ideas though? [03:49:17] <twincest> you could try enabling the cache manually with format, i don't know if zfs will override it again [03:49:28] <Symmetria> heh twincest see msg again btw when u get a chance [03:49:48] *** packetjunkie has quit IRC [03:50:25] <BR_> twincest: okies, checking [03:51:24] <jteo> BR_: you added whole LUNS, and not slices? [03:51:35] <BR_> mind if I paste 4 lines? [03:51:54] <BR_> or want it in a pastebin? [03:52:12] <jteo> pastebin. [03:54:12] <BR_> http://pastebin.ca/225654 [03:54:44] <BR_> jteo: yes whole luns are added from the SAN [03:54:53] *** Griff has quit IRC [03:56:57] <jteo> BR_: zfs uses the caches, but it forces a write cache flush when it commits. [03:57:06] <jteo> BR_: your array caches are non-volatile? [03:57:29] <BR_> Battery Backed up, but just normal SDRAM [03:57:38] <jteo> BR_: ah. semi-volatile. [03:57:39] <jteo> ;) [03:57:52] <jteo> BR_: there's an RFE to not flush the caches i think. [03:58:06] <richlowe> I thought it was to more selectively flush them. [03:58:21] <BR_> jteo: excuse my ignorance, but RFE? [03:58:24] <richlowe> If we're thinking of the same thread, there was then discussion about gathering a list of 'broken' devices? [03:58:28] <richlowe> BR_: Request For Enhancement [03:58:35] <BR_> richlowe: ah okies... [03:58:43] <jteo> richlowe: i vaguely recall the details. but if the cache is battery backed, then flushing won't be neccessary yes? [03:58:56] *** B|nTaRa has quit IRC [03:59:31] <BR_> jteo: not so much sure as 'not neccessary' but not needed more like... I can see the possibility to data coruption with the ZFS way of thinking with the whole transactional commits [03:59:56] <gisburn> why are the flushes needed at all ? zfs stores the checksums for the data and should be able to figure out which data are valid and which are damaged... [04:00:26] <BR_> jteo: of the cache is re-organised by the array in to such a way that part of the transaction is written and other parts not at the time of a power fail then I can see a possibility to have a bad commit. [04:01:26] <BR_> and as I understand the ZFS way of things, thats a bad thing(TM) [04:02:41] <jteo> BR_: that's why i asked whether if it was non-volatile. [04:02:58] <BR_> unfortunatly I am now in the same boat as Symmetria, I will get dumped by my GF if I don't sleep now ;) I will catch up with the scrollback tomorrow morning, thanks for your answered though guys (and any more that come up) [04:04:59] <richlowe> gisburn: the idea is to stop data geting damaged, not just to notice it after the fact... [04:10:56] *** Risky has joined #opensolaris [04:16:47] *** drio has quit IRC [04:19:06] <spawrq> sun needs to bring the prices of the sparc's down [04:19:10] <spawrq> waaaaay down. [04:19:28] <gisburn> spawrq: yes [04:19:38] <spawrq> maybe they'd actually be competitive. [04:19:43] <richlowe> t1k is around $3,000, isn't it? [04:19:50] <Auralis> that sadly requieres them to sell a lot more of them in the first place [04:19:56] <richlowe> most of the other stuff is oddly priced though. [04:20:03] <richlowe> especially since most of it is less desirable than a T1 :) [04:20:09] <spawrq> yep. [04:20:30] <spawrq> excellent, killer, fantastic .... but nobody can justify the cost. [04:20:37] <spawrq> well [04:20:49] <richlowe> I don't think you'd get anything comparible to the t1000 for a better price. [04:20:58] <spawrq> all i'm saying is if they cut the prices, i'll bet they'd sell a boatload more. [04:21:02] <richlowe> most of the other machines though... [04:22:01] <gisburn> richlowe: my point is that Sun has NO offering for developers which is below this point. And $3000 is completly off limits for both students and universities until a project has speific demand for this type of machine. [04:22:16] <jteo> i agree with gisburn. [04:22:27] <spawrq> i mean [04:22:28] <jteo> sparc wise obviously. [04:22:36] <spawrq> why am i stuck with a crappy IIIi for my workstation? [04:22:44] <gisburn> richlowe: Sun could easily bulid a UltraSPARC-IIe-based machine for less than $500 [04:22:45] <spawrq> might as well just go get an opteron. [04:22:58] <richlowe> gisburn: Probably. [04:23:01] <richlowe> gisburn: and it'd suck.- [04:23:04] <richlowe> gisburn: utterly. [04:23:07] <delewis> or you could just buy a used SB1000 [04:23:08] <Auralis> yuck, us2e ... [04:23:09] <delewis> for the same price [04:23:16] <delewis> that blows the UltraSPARC-IIe away [04:23:20] <richlowe> they sold the b150 for $1,000-ish, didn't they? [04:23:21] <richlowe> and it sucked. [04:23:22] <richlowe> utterly. [04:23:25] <richlowe> so... [04:23:26] <delewis> richlowe: yeah. [04:23:33] <delewis> SB100 and SB150 were both in the $1k range. [04:23:39] <spawrq> yah [04:23:44] <gisburn> richlowe: the Blade150 sucked because if sucky disks and sucks PCi controllers. [04:23:45] <delewis> of course, they wouldn't have sucked *that* bad if Sun chose a decent IDE chipset [04:23:46] <Auralis> yes, the entire point of the sb100 was to have a sub 1000 price [04:23:49] <delewis> *instead* [04:23:51] <delewis> they chose cmdk [04:23:56] <delewis> with *no* DMA whatsoever [04:24:03] <gisburn> yeah [04:24:08] <delewis> and the fact it had a unified PCI bus didn't help matters, either. [04:24:11] <Auralis> and don't forget that single pci bus for the entire machine [04:24:19] <spawrq> the 100/150's were smellarooni [04:24:19] <delewis> Auralis: :-) [04:24:27] <richlowe> I'd rather they price themselves out of a lot of markets, than release another cheap workstation people are embarrassed to mention, I guess. [04:24:36] <richlowe> u5, u10, b1[05]0, are enough of those. [04:25:00] <spawrq> we've still got some u2's hanging around. [04:25:04] <delewis> SB1000s weren't really that badly priced by comparison, either for the time [04:25:11] <richlowe> spawrq: the u2 is better than any of those I just listed. :) [04:25:14] *** gleaken has quit IRC [04:25:19] <richlowe> spawrq: that is, in fact, the problem with the ones I mentioned :) [04:25:19] <delewis> $10k entry-level isn't that bad for a UNIX workstation, considering at the time SGI was selling Octanes for about $30-$40k [04:25:33] <delewis> I'd rather save up for a $10k workstation that buy a $1k-$3k POS. [04:25:34] <spawrq> richlowe- but we also have a boatload of 10's and 60's. [04:25:37] <spawrq> and a bunch of 5's as well. [04:25:55] <Auralis> well, at the time the sb1000 was released, it was infact speedwise on par with the x86 crowd [04:26:00] <delewis> Auralis: yes [04:26:05] <spawrq> i use a 10 for a footstool at work. [04:26:13] <delewis> and it was a very economical workstation despite how much power it had [04:26:14] <elektronkind> the b100 makes for a great.... print server... [04:26:23] <delewis> HP and SGI both had *really*, *really* overpriced offerings at the time [04:26:37] <delewis> a J2240 (an HP9000 that compared to the SB1000) at the time was about $65k [04:26:45] <richlowe> delewis: last I knew, SGI still do. [04:26:59] <delewis> richlowe: :-) [04:27:06] <delewis> their prices have come down [04:27:09] <delewis> (sort of) [04:27:11] <richlowe> even their "we're trying to ditch all our mips stock and get out of chapt. 11" sale doesn't bring them even close to far enough down. [04:27:14] <elektronkind> speaking of the blah workstations, I was really bummed about the X1 and its follow-on, the V100 [04:27:15] <delewis> their Itanium workstation has a $10k-$30k price range [04:27:25] <delewis> rather than $30k-$50k like the Octane did [04:27:50] <gisburn> Note that the Ultra5 wasn't bad when it was used with a Symbios SCSI controller, 512MB RAM and a non-Sun m64 compatible card. [04:27:51] <richlowe> elektronkind: that's kinda my point in all this. [04:27:51] <delewis> though, the Octane was *far* more innovative than their Itanium workstation [04:27:56] <richlowe> what use is cheap if the machines end up sucking? [04:27:56] <Auralis> i liked the X1, it was a ok little box, just hellish loud and i sucked that the second hd carrier was not included [04:27:59] <delewis> XIO rocked :-) [04:28:12] <delewis> nothing like a 3.2Gbps bus in 1998. [04:28:27] <elektronkind> yeah, XIO rocked. I loved my o2000 [04:28:47] <elektronkind> quad fast ethernet card where all ports could get line speed.. in 1998. [04:29:06] <delewis> regardless, I doubt we're ever going to see cheap, decent SPARC systems from Sun, anyway. [04:29:26] <delewis> cheap goes against everything UNIX workstations stand for -- quality and elegance. [04:29:27] <gisburn> delewis: yeah [04:29:36] <gisburn> er [04:29:38] <gisburn> no. [04:29:52] <gisburn> delewis: I think a cheap development box would be a great thing. [04:30:05] <Auralis> that was the reason behind the sb100 [04:30:17] <Auralis> give devs a cheap 64bit box [04:30:18] <delewis> which cut way too many corners [04:30:22] <delewis> in order to make the $1k price [04:30:41] <delewis> and regardless [04:30:45] <gisburn> Auralis: grrr... the blade100 was completely f*cked.up because of wrong decisions and bad drivers. [04:30:46] <delewis> look at the current SPARC offerings [04:30:47] <delewis> with the Ultra 25 [04:30:52] <delewis> things *have* gotten much better [04:30:57] <gisburn> Auralis: Linux on a Ultra5 has a faster filesystem than Solaris 11 [04:31:03] <delewis> $3k for an entry-level SPARC with an UltraSPARC-IIIi and SCSI? [04:31:06] <gisburn> MUCH faster [04:31:09] <delewis> that's a *huge* improvement. [04:31:39] <gisburn> delewis: as I said $3k is off-limits for normal developers these days. [04:31:54] <gisburn> at least in the opensource field [04:32:01] <Auralis> now, if someone could put back together a nice pizzabox workstation [04:32:05] <delewis> gisburn: but it is a imrpovement, though. [04:32:18] <delewis> improvement, rather. [04:32:33] <delewis> you could always get a refurb [04:32:36] <gisburn> delewis: correct me if I am wrong: one of the sunrays had an ultrasparc CPU, right ? [04:32:39] <delewis> SB150 from Sun Store :-) [04:32:47] <delewis> gisburn: I highly, highly doubt that. [04:32:59] <delewis> the Javastations used a sun4m [04:33:02] <Auralis> sunrays are 110mhz microsparcs [04:33:05] <delewis> but that wasn't Ultra [04:33:53] <delewis> gisburn: if you really want to develop on an UltraSPARC system, the used market is fairly good, nowadays. [04:34:08] <delewis> SB1000s are *way*, *way* under the $1k range [04:34:17] <gisburn> delewis: not here in germany. [04:34:20] <Auralis> sadly so, at times as low as 300 bucks [04:34:23] <richlowe> the older sunrays had real sun4m procs, I believe. [04:34:29] <richlowe> the newer ones lack an MMU [04:34:32] <gisburn> delewis: I tried to aquire one on ebay and gave up at 4800 euro [04:34:33] <delewis> and if you want to buy it from someone reputable that'll give it a warranty, Anysystem.com sells a base SB1000 for about $800 USD [04:34:45] <richlowe> that's from memory of something someone said in here though (gdamore?) [04:35:23] <richlowe> gisburn: that can't have been recently. [04:35:24] <delewis> gisburn: anysystem.com will ship to Germany. [04:35:36] <richlowe> or you picked a real real bad day. [04:35:37] <delewis> how much for, I'm not sure, though [04:35:42] <elektronkind> 100 MHz microSPARC IIep System Board. Includes ATI 128 Ultra graphics. [04:35:45] <richlowe> delewis: Not Cheap [04:35:49] <elektronkind> that's the Sun Ray 1's specs [04:35:50] <richlowe> b1000 ways around 80lb [04:35:50] *** Dr_Jekyll has joined #opensolaris [04:35:55] <delewis> richlowe: true [04:36:04] <richlowe> heh, "weighs", rather. [04:36:11] <delewis> I know because I have sitting beside me, need I not be reminded. :-) [04:36:40] <elektronkind> so the Sun Ray 1 had a 100Mhz MicroSPARC IIep... basically a souped-up SPARCstation LX [04:36:52] <delewis> more like a SPARCstation 5 [04:37:27] <elektronkind> that used a TurboSPARC chip iirc [04:37:34] <delewis> not all of them [04:37:39] <gisburn> HyperSPARC [04:37:55] <Auralis> ss5 was microsparc and the 170er turbosparc [04:37:58] <delewis> the low-end ones had TurbSPARCs and the high-end ones had HyperSPARCs from Fujitsu [04:38:01] <delewis> which were the 170Mhz variants [04:38:03] <Auralis> hypersparc was ss10/20 only [04:38:10] <elektronkind> HyperSPARC was only made by Ross, Inc, for the SPARCstation 10, 20, and the sun4d boxes [04:38:10] *** Dr_Jekyl1 has quit IRC [04:38:11] <delewis> TI shipped 75 and 110MHz variants [04:38:42] * gisburn mourns the days of the sun4d [04:38:52] <elektronkind> Tatung made a 85Mhz sparcstation 5 clone, too. I think Axil also made one with the same specs [04:39:04] <delewis> maybe I'm thinking about the 85 then [04:39:15] <delewis> I know there's a 110, because I have one sitting in my closet [04:39:26] <Auralis> thats a microsparc [04:39:44] <gisburn> Ok, little quiz: [04:39:45] <gisburn> "I'm struck and cut, shaped and cooled, then bound by rings to release whats stored." - what am I ? [04:39:46] *** polk__ has joined #opensolaris [04:40:03] <Auralis> a key [04:40:16] <gisburn> heh [04:40:22] <delewis> hey I was close :-) [04:40:29] <delewis> I knew it had something to do with metallurgy. [04:40:32] <gisburn> Auralis: I asked this earlier today and noone except stevel figured it out. [04:40:36] <delewis> (I guessed coinage earlier) [04:43:15] <delewis> D1000 is still supported, is it not? [04:43:24] <delewis> SunSolve only says up to Solaris 9 supports it [04:43:29] <gisburn> http://www.opensolaris.org/os/project/ksh93-integration/ - resouce not found ? [04:43:31] <delewis> (which I know is true for the A1000) [04:43:32] <gisburn> HEEEELLLLP [04:43:37] <delewis> gisburn: it's been that way for days.. [04:43:40] * delewis thought you knew [04:44:04] <gisburn> delewis: no [04:44:09] <gisburn> delewis: what happened ? [04:44:34] <delewis> gisburn: I think it disappeared Tuesday or Wednesday [04:46:08] <Auralis> well, a d1000 is just a dual pathed jbod, there is nothign requiering special drivers to it [04:46:19] <delewis> I didn't think that was the case [04:46:34] <elektronkind> delewis: I bet the monitoring software for it doesn't work on s10, but naturally it's just a dumb differential scsi device [04:46:41] <delewis> elektronkind: right [04:46:55] <elektronkind> it was only environmental monitoring, though [04:46:58] <delewis> and there's really no point for the monitoring system, anyway -- unlike the A1000 is doesn't do RAID [04:47:01] <elektronkind> nothing confi-wise [04:47:01] <delewis> ah [04:47:36] <elektronkind> the only thing between the scsi ports on the back and the disks in the front on a D1000 is a differential to single-ended conversion chip [04:48:05] <elektronkind> and a few thermoresistors [04:48:06] *** jmcneill has joined #opensolaris [04:48:25] *** nwf has quit IRC [04:50:03] <gisburn> delewis: please report such problems ASAP next time... please... Ok ? [04:50:10] <delewis> gisburn: sure [04:50:23] <gnu2it2> is ther a moniter or top app on freash install or do i need to find it elsewhere? [04:50:31] <delewis> gnu2it2: prstat [04:50:34] <delewis> which is superior to top [04:51:05] <gnu2it2> u rock, thanks [04:53:11] *** Gman has quit IRC [04:55:20] <dwc-> prstat needs the status line at the top of top [04:55:33] <jteo> file an RFE.;) [04:55:41] <dwc-> I think someone did [04:55:54] <dwc-> one of the other previous 982375932 times this has been mentioned here ;) [04:57:00] *** OnkelSchorsch has joined #opensolaris [04:58:53] <gnu2it2> is there an easy way to move from sol 10 to nv current without reinstall ? [04:59:41] <jteo> gnu2it2: to an upgrade. [04:59:43] <jteo> *do [04:59:50] <gnu2it2> yes [05:02:25] *** Gman has joined #opensolaris [05:07:18] *** swa has joined #opensolaris [05:09:20] <Sieghard> what svn tag is 10u2 again? [05:09:34] <delewis> it does not have one [05:09:36] <richlowe> none of them. [05:09:37] <jteo> s10 is a different tree. [05:09:43] <richlowe> it's an entirely different source tree, and not open. [05:09:49] <Sieghard> oh how about the zfs that made it to 10u2 [05:11:31] <Sieghard> wow searching for that on google returned this channel [05:14:33] <twincest> that was snv_27 originally with some patches from later versions [05:14:43] <twincest> (actually quite a few) [05:14:44] <Sieghard> ok thanks [05:15:20] <richlowe> As I understand it, there's no "Take something as of mktg+1 and merge all of it back to mktg" to these things. [05:17:35] <elektronkind> I am damn ready for u3 to come out [05:18:00] <elektronkind> just the zfs patches alone are worth it [05:20:15] *** axisys has joined #opensolaris [05:22:09] *** jsubl2 has quit IRC [05:23:21] <myrkraverk> what can I use to list usb-devices? [05:23:30] <delewis> cfgadm -l [05:24:21] <myrkraverk> k [05:24:46] <Sieghard> that shows all usb devices? [05:24:53] <Sieghard> why is there sata and usb together [05:24:54] <myrkraverk> can I get a list of all serial controllers/devices (UART?) [05:25:47] <delewis> Sieghard: it shows all dynamically reconfigurable resources [05:25:56] <delewis> which includes devices that support hot-swap. [05:26:00] <Sieghard> delewis: ah, so yay for sata framework [05:26:17] <delewis> SCSI and fibre-channel show up, too :-) [05:26:28] <Sieghard> delewis: so why not firewire? i don't see firewire on the list [05:27:56] <delewis> Sieghard: not sure [05:28:05] <delewis> but I did find a type in the ieee1394(7D) man page [05:28:15] <delewis> typo* [05:28:24] <Sieghard> that's a good find [05:28:27] <myrkraverk> I've plugged in my usb-serial thingy - and I think it's configured - what do I use to see the RS-232 controller? [05:28:28] <delewis> "For more information on USB, refer to the 1394 Trade Association web-site at http://www.1394ta.org." [05:28:45] <Sieghard> what on earth [05:28:52] <elektronkind> cfgadm is for managing any type of removable hardware device on your system [05:29:04] <richlowe> any type which supports the right interfaces. [05:29:07] <elektronkind> this means anything from USB devices to CPU and memory boards [05:29:30] <richlowe> ls /usr/lib/cfgadm/ [05:29:32] <elektronkind> right (there has to be a specific cfgadm module for it, but that's generally not a problem) [05:29:44] *** jerome__ has left #opensolaris [05:29:49] <richlowe> elektronkind: just bridging it back to the firewire discussion :) [05:29:54] <elektronkind> ok :) [05:30:29] <Sieghard> firewire should go on the list then [05:30:46] <Sieghard> doesn't it support the right interfaces? [05:30:51] <Sieghard> it's very hot-pluggable [05:30:58] <richlowe> I don't see a cfgadm module. [05:31:05] *** jmcneill has left #opensolaris [05:31:05] <richlowe> but I don't have any devices I could plug in, to see if anything changes. [05:31:10] <Sieghard> maybe it's too versatile to be included, because firewire can also be ethernet [05:31:19] <elektronkind> so can usb [05:31:26] <Sieghard> can it? [05:31:35] <richlowe> firewire can't "be" eithernet. [05:31:41] <Sieghard> i don't see any entries for usb in ifconfig [05:31:44] <Sieghard> but i do for firewire [05:31:48] <richlowe> but you can push IP over it. [05:32:06] <elektronkind> Sieghard: ah, I thought you were referring to adaptors [05:32:20] <richlowe> (eithernet is quite the typo.) [05:32:25] <Sieghard> richlowe: well that's essentially what i meant [05:32:29] <richlowe> anyone need a new compatibility slogan? :) [05:32:31] <Sieghard> ok [05:32:45] <Sieghard> let's call it tcp over firewire [05:32:49] <Sieghard> tcp/ip [05:34:28] <Sieghard> well i guess you can push udp over it as well [05:34:39] <elektronkind> it's IP over firewire [05:34:51] <gisburn> cool [05:34:55] <Sieghard> ipv6 [05:34:58] <gisburn> elektronkind: I bet it is IPv4-only [05:34:59] <elektronkind> ieee 1394 is the link layer [05:35:13] <Sieghard> gisburn: i've used ipv6 with it [05:35:23] <gisburn> Sieghard: ah... ok... [05:35:52] <myrkraverk> can I use tip for PPP tcp/ip - or do I use something else? [05:36:09] <elektronkind> myrkraverk: pppd ? [05:36:10] <Auralis> there is pppd [05:36:52] <myrkraverk> ah, k [05:37:08] <richlowe> if you can do it with tip, I for one would be impressed :) [05:37:19] <myrkraverk> ;) [05:37:22] <richlowe> ... where's nrubsig, he could do it in ksh :) [05:37:27] <myrkraverk> haha ;) [05:37:28] *** jsubl2 has joined #opensolaris [05:37:40] <myrkraverk> do I use pppd on the other host as well? [05:37:56] <elektronkind> you would [05:38:00] <myrkraverk> ok [05:38:03] <Sieghard> ok for some reason i now have a gif0 and stf0 with mtu 1280 in my list of network devices [05:38:04] <elektronkind> pppd is for IP over a serial link [05:38:05] <Sieghard> no idea what they are [05:38:23] <Sieghard> is someone trying to connect an external modem to a serial port? [05:38:52] <myrkraverk> elektronkind: yes - since I can't use e1000g for now - I thing I'll give my usb<--->RS-232 with a null-modem a try ;) [05:39:07] <myrkraverk> Sieghard: ^^^^^ [05:39:16] <Sieghard> ahh [05:39:23] <Sieghard> i guess it's pretty fast [05:39:25] <elektronkind> can't use your e1000g interface why? [05:39:26] <Sieghard> usb 2.0? [05:39:42] <gisburn> where is the jive page for website-discuss@ ? [05:39:57] <richlowe> gisburn: hit opensolaris.org/os/discussions, and get to it that way. [05:40:05] <richlowe> the jive urls are human-proof. [05:40:10] <myrkraverk> elektronkind: some bug in the driver it appears; same as: http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6431317 but different card [05:40:29] <myrkraverk> (I have filed a bug report) [05:40:45] <elektronkind> that sucks [05:40:49] <myrkraverk> yes [05:41:22] <elektronkind> it's probably some stupid chip rev check in the driver code itself [05:41:27] <myrkraverk> yes [05:42:02] <elektronkind> I hate those kinds of bugs :P or rather, running into them myself. [05:42:17] <elektronkind> had that problem on a few scsi cards and the glm driver [05:42:27] <myrkraverk> I may give it a try to port the (some-or-other)-BSD em driver from intel - but that will probably take me a bit of a time, so having the serial in the mean time will help [05:42:28] <richlowe> Hm. [05:43:27] <elektronkind> myrkraverk: alternative is to get a cheap relatek card :) I had to do that while waiting for my rev of the nvidia driver (nge) to support the nforce 410 chipset [05:43:50] <richlowe> murayama has an 'em' driver, that may work. [05:44:23] *** jsubl2 has quit IRC [05:44:28] <richlowe> I wonder if they're going to integrate murayama's drivers that would replace closed ones. [05:44:44] <richlowe> (rather than just additional ones, which I think someone or other is working on) [05:45:45] <myrkraverk> richlowe: tried it already [05:46:16] <myrkraverk> elektronkind: I'm on a laptop - that limits my "alternatives" a bit [05:49:45] <elektronkind> bummer [05:52:05] <myrkraverk> well, I'm going to sleep now - I hope to get some network tomorrow ;) [05:52:12] <Sieghard> myrkraverk: does it have a cardbus slot? [05:52:43] <Sieghard> i'm looking for e1000g cardbus cards that support jumbo frames [05:52:45] <myrkraverk> Sieghard: yes, I think so - but I'm not sure my controller is supported [05:53:22] <Sieghard> my current e1000g doesn't support jumbo frames, and basically the only computer that doesn't in the network [05:55:11] *** paxc has joined #opensolaris [05:55:16] <myrkraverk> hmm, is xpdf not part of EXCR? [05:55:25] <delewis> myrkraverk: no [05:55:28] <delewis> but evince is [05:55:28] <richlowe> evince is. [05:55:34] <delewis> and acroread ships on Solaris/SPARC [05:55:50] *** gisburn has left #opensolaris [05:56:01] <richlowe> great, he's using that joke again. [05:56:18] <myrkraverk> what joke? [05:56:58] <delewis> that's just payback for our ksh93 puns [05:57:27] <richlowe> I'd be in a way better mood if I could figure out this X crash. [05:57:32] <richlowe> at least to the degree I could stop making it happen. [05:57:33] <myrkraverk> anyways, I'm off; night all [06:00:07] *** paxc has quit IRC [06:07:39] <dwc-> I'll take xpdf over evince and acroread [06:07:52] <richlowe> I thought xpdf and evince used the same pdf stuff? [06:08:34] *** dclarke_ has joined #opensolaris [06:09:41] <delewis> both make use of a ghostscript back-end, IIRC. [06:10:04] *** dclarke_ is now known as d_m_c [06:10:11] <d_m_c> test ... [06:10:16] <d_m_c> dennis here [06:10:22] <d_m_c> on a different machine [06:10:31] <d_m_c> windows 98 in fact .. gack [06:11:20] * d_m_c patches his V120 [06:11:55] *** jsubl2 has joined #opensolaris [06:17:02] *** Triskelios has quit IRC [06:17:36] *** sparkleytone has quit IRC [06:17:42] *** Triskelios has joined #opensolaris [06:17:47] <gnu2it2> how good is a sunblade 2000 w 2 x 900mgh proc, 2gb ram and a 36 gb hd? worth $400 ? [06:17:51] *** xzilla has quit IRC [06:17:53] *** xzilla has joined #opensolaris [06:17:56] <d_m_c> yes! [06:17:59] <delewis> gnu2it2: definitely. [06:18:02] <d_m_c> worth it big time [06:18:16] <delewis> what's the framebuffer? [06:18:27] <delewis> pgx32 or elite3d, I'm guessing? [06:18:29] *** sparkleytone has joined #opensolaris [06:18:56] <gnu2it2> Creator 3D or Elite 3D [06:19:16] <delewis> take the Creator if you want to run at greater than 1280x1024 [06:19:29] <delewis> sacrificing 3D in the process [06:19:34] *** vipe has quit IRC [06:19:38] <delewis> (actually, the Elite3D sucks, too) [06:19:39] <d_m_c> its a great framebuffer [06:19:46] <d_m_c> not the e3d [06:19:49] <gnu2it2> those FC-AL disk are $$$ but specs are nice [06:19:58] <delewis> gnu2it2: not really [06:20:01] <delewis> FC-AL is dirt-cheap [06:20:07] <delewis> cheaper than Ultra160 SCSI in almost all cases. [06:20:08] <d_m_c> funny bug : 6471607 gethostbyaddr() & friends are overly paranoid [06:20:08] *** vipe has joined #opensolaris [06:21:07] <gnu2it2> buck a gb up to 36,, then climbs from what i can tell [06:21:45] <delewis> gnu2it2: compared to IDE or SATA, I suppose it would be much more expensive, but in the realm of SCSI, not really. [06:22:37] <gnu2it2> bummer is i only have 1 sun monitor that supports 13w3,, and it's guns are failing [06:22:54] <delewis> get yourself a 13w3->VGA adapter. [06:22:59] <gnu2it2> screen is kinda blue [06:24:14] <gnu2it2> will play server/sandbox mostly so after setup really doesnt need vid [06:24:38] <delewis> it doesn't *need* video, anyway. [06:24:45] <delewis> during any point of its operation. [06:25:58] <gnu2it2> delewis: u support commercial accounts ? [06:26:12] <delewis> gnu2it2: nope [06:26:50] <gnu2it2> had to chuckle other day, [06:27:28] <gnu2it2> made comment about headless servers, ,and they didnt understand [06:28:08] <Auralis> x86 whinnies [06:28:20] <gnu2it2> hehe [06:28:26] <d_m_c> ha [06:29:33] <gnu2it2> 1/2 were ex novell/win shops [06:30:08] <delewis> and there other half were Lusers? [06:31:31] <gnu2it2> wow,, wish i had that thought the other day. would have really alianated me [06:37:28] <Gadzooks> d_m_c < can we pkgrm CSWmozilla off of ra/thor? [06:37:35] <Gadzooks> :D [06:38:14] <d_m_c> oh heck [06:38:30] <d_m_c> yeah [06:38:31] <d_m_c> any idea why ? [06:38:47] <Gadzooks> yelp/evolution [06:39:38] <Gadzooks> (replaced with firefox/seamonkey) [06:39:43] <d_m_c> fixable stuff ? [06:39:52] <d_m_c> well .. I prefer seamonkey anyways [06:39:59] <d_m_c> okay .. let me look into that [06:40:17] <d_m_c> I'm sort of stuff watching osiris patch [06:40:25] <Gadzooks> ok [06:41:18] <Gadzooks> evolution builds...just making sure its not linking to CSWmozilla and causing me grief. [06:41:24] <d_m_c> okay [06:41:46] <d_m_c> can you login to login ? does it seem to hang there for a minute ? [06:42:07] <d_m_c> for me it seems ot hang as if it can not reverse lookup where I am coming from [06:42:54] <Gadzooks> I'll retry [06:43:40] <Gadzooks> Last login: Fri Oct 27 19:54:54 2006 from {hangs here}... [06:43:41] <d_m_c> ra is doing the same dman thing [06:43:50] <d_m_c> weird [06:43:59] <d_m_c> I have a new problem on my hands here [06:44:05] <Gadzooks> NFS server core-hme1 not responding still trying [06:44:14] <d_m_c> oh hell [06:44:17] <Gadzooks> oh-oh [06:44:21] <d_m_c> oh bloody hell [06:44:31] <d_m_c> not the NFS server ... [06:44:40] <d_m_c> say it isn't so [06:45:08] <d_m_c> Houston we have a problem [06:45:54] <d_m_c> can you do me a favour and drop a message on the maintainers list that tells eveyone we are aware of the problem and I am looking into it ? [06:46:02] <Gadzooks> ok [06:46:25] <d_m_c> looks like the NFS server has had a rough night tonight [06:46:45] <Gadzooks> ok..I'm on login as well...just starting doing this..was fine until recently. [06:47:22] <d_m_c> yeah .. looks to be .. some new problem [06:47:29] <d_m_c> must have happend minutes ago [06:47:32] <Gadzooks> 'This machine is NOT in a good state.' [06:47:37] <d_m_c> something we did ? [06:47:47] <d_m_c> thats an old message .. I better fix that [06:47:58] <d_m_c> its not in a bad state [06:48:03] <d_m_c> just a little flaky [06:48:22] <d_m_c> actually login is damn stable [06:48:33] <d_m_c> but I want to replace it with Sol10u2 [06:48:47] <d_m_c> fully patched and ipfilter yada yada yada [06:48:56] *** mr_goose has quit IRC [06:49:17] <d_m_c> NFS server core not responding still trying [06:49:36] <d_m_c> core-hme1:/export/nfs/share just came back online [06:49:39] <d_m_c> wtf ? [06:50:03] <richlowe> Gman: if you added me to that interest list, I never got notified of it. [06:51:15] <d_m_c> yikes [06:51:16] <richlowe> Gman: though if you didn't, that's fine. Just wondering if the whole thing is acting up [06:51:30] <d_m_c> Oct 28 00:49:12 core ufs: [ID 213553 kern.notice] NOTICE: realloccg /export/home: file system full [06:51:49] *** jsubl2 has quit IRC [06:52:10] <d_m_c> core dump ate the whole filesystem [06:52:20] <richlowe> what dumped? [06:52:25] <d_m_c> Gadzooks: wow [06:52:44] <d_m_c> screen-4.0.2 [06:52:46] <richlowe> generally takes infinite recursion and a lot of it to do that, unless you have some immense process. [06:52:58] <delewis> screen-4.0.2 dumped and filled up all of /export? [06:52:58] <richlowe> d_m_c: pstack it, tell me how long it takes :) [06:53:12] <d_m_c> you want me to run pstack on that ? [06:53:19] <d_m_c> can I move it first ? [06:53:20] <richlowe> for some reason I get a kick out of core dumps that take half an hour to stack trace :) [06:53:22] <Gadzooks> core dump? just a simple stagepkg of evolution..worked fine before... [06:53:27] <d_m_c> to somewhere .. anywhere .. else [06:53:38] <Gadzooks> odd [06:53:48] <d_m_c> well something dumped core and ate the whole bloody filesystem man [06:53:54] <d_m_c> and it happened as we were working [06:53:55] <Gadzooks> eek [06:54:08] <d_m_c> and even weirder ... it was screen that dumped [06:54:12] <richlowe> d_m_c: how big? [06:54:16] <d_m_c> screen is pretty damn stable stuff [06:54:27] <delewis> there's no way screen would fill the filesystem with a core dump [06:54:28] <Gadzooks> only a 1-2MB [06:54:29] <delewis> it's *tiny* [06:54:42] <Gadzooks> let me try again... [06:54:48] <d_m_c> 166310090752 bytes [06:54:52] <richlowe> delewis: infinite recursion, unlimited dump size, unlimited stack size. [06:55:10] <Gadzooks> my my my [06:55:17] <d_m_c> yeah man [06:55:17] <richlowe> delewis: careless use of irssi's perl interface will get you multi-hundred meg irssi core's. Just recurse in their until you run out of stack :) [06:55:22] <d_m_c> it killed the NFS server [06:55:37] <delewis> richlowe: which is why I don't bother building irssi with Perl support :-) [06:55:51] <d_m_c> sorry .. let me rephrase that ... nothing kills Solaris 10 [06:55:56] <d_m_c> it became insanely busy [06:56:08] <d_m_c> too busy to perform NFS tasks at least [06:56:19] <d_m_c> $ uptime [06:56:20] <d_m_c> 12:55am up 101 day(s), 5:51, 3 users, load average: 0.06, 3.36, 3.37 [06:56:21] <d_m_c> $ [06:57:05] <d_m_c> Oct 28 00:02:20 core nfssrv: [ID 694464 kern.warning] WARNING: nfsauth upcall failed: RPC: Operation in progress [06:57:06] <d_m_c> Oct 28 00:49:08 core ufs: [ID 845546 kern.notice] NOTICE: alloc: /export/home: file system full [06:57:07] <d_m_c> Oct 28 00:49:12 core ufs: [ID 213553 kern.notice] NOTICE: realloccg /export/home: file system full [06:57:08] <Gadzooks> I knew Gnome sucked up resources..but that is plain silly! ;> [06:57:26] <d_m_c> let me check to see what this pstack thing says [06:57:44] <d_m_c> richlowe: you really want to know ? [06:57:59] <richlowe> d_m_c: I want to know how long it takes, not what it says :) [06:58:08] <richlowe> I expect what it says will be, well, voluminous :) [06:58:27] <d_m_c> so then .. time -p pstack core > /dev/null ? [07:00:59] <d_m_c> Gadzooks: I think we have our NFS server back .. I think that something really really "bad" took place with a screen session or at least screen task [07:01:24] <d_m_c> it ate all avaliable memory and swap I guess and then dumped core .. forever [07:01:45] <d_m_c> however .. and this really bugs me .. why would a core file be larger than RAM ? [07:01:52] <Gman> richlowe, sorry, have my gf's nieces over at the moment, so a little distracted [07:01:58] <Gman> will cc' you on the interest list now [07:02:11] <richlowe> Gman: nah, I just wondered if it was broken, or if you hadn't done it. [07:02:19] <richlowe> Gman: it's not like anyone will add to it before monday or so, anyway :) [07:02:42] <Gman> done [07:02:43] * richlowe tries to keep track of what b.o.o is doing wrong. [07:02:44] <richlowe> foolish, I know. [07:02:49] <Gman> yeah, i saw your comments ;) [07:03:10] <Gman> '[this field filled because b.o.o second guesses humans, badly]' [07:03:13] <d_m_c> oops .. [07:03:20] <richlowe> well it does! [07:03:28] <Gadzooks> d_m_c.. seems ok at the moment... [07:03:37] <richlowe> Gman: *you* find actual result/expected result/etc, for that. [07:03:40] <d_m_c> yeah .. smooth recover eh ? [07:03:43] <richlowe> Gman: and b.o.o *forces* you to fill those fields :) [07:03:46] <Gman> richlowe, heh, most of that stuff gets cut and pasted into the description [07:03:49] <richlowe> Gman: so I fill them with sarcasm :) [07:03:50] <Gman> ahh, silly [07:04:04] <Gman> did you log a bug? :) [07:04:59] <gnu2it2> what differs on FC-AL interface ? i see different part# for 36.4 drives? [07:05:00] <jteo> .... [07:05:09] <Gadzooks> d_m_c...well...lets give it a few minutes... [07:05:40] <d_m_c> lets give it a moment ot catch its breath I guess [07:05:44] *** sahafeez has joined #opensolaris [07:05:51] <d_m_c> Removal of <CSWmozilla> was successful. [07:06:17] <sahafeez> do the volume keys on sparc system only work on the internal speaker? [07:06:18] <d_m_c> a lot of user sessions got dropped also [07:06:38] <delewis> sahafeez: no [07:06:47] <delewis> they effect whatever you're using as your output sound device [07:07:00] <sahafeez> well, then they are borked. [07:07:07] <richlowe> Gman: I logged one about it mandating them on RFE's. [07:07:08] <delewis> sahafeez: the sound applet is broken in JDS [07:07:17] <sahafeez> well, there you go [07:07:25] <delewis> works in CDE, though.. [07:07:27] <richlowe> Gman: on bugs, it makes some sense, kinda, except they're not applicable to all bugs. [07:07:32] * delewis ducks [07:07:32] <delewis> :-) [07:07:33] <d_m_c> works on my machine [07:07:41] <d_m_c> on snv_46 [07:07:56] <delewis> d_m_c: not in b48 or b50 [07:08:03] <d_m_c> odd [07:08:06] <sahafeez> well i get the popup but no workie [07:08:10] <d_m_c> well .. I'm lucky then [07:08:11] <delewis> sahafeez: right [07:08:40] <richlowe> Gman: 6459322 [07:08:52] <richlowe> Gman: I especially like it requiring the steps to reproduce an RFE :) [07:09:06] *** lplatypus has quit IRC [07:09:16] <d_m_c> Gadzooks: Removal of <CSWmozilla> was successful. [07:09:27] <Gadzooks> thanks [07:09:57] <sahafeez> strange question but....vfstab did not get replaced by somthing in sol10 right? if i want to add a nfs mount to mount at boot that is the best place to do it [07:10:08] <d_m_c> note to self : don't ever try to remove mozilla .. it kills the NFS server [07:10:11] <delewis> sahafeez: vfstab [07:10:24] <Gadzooks> I built the same package without NFS crashing..we seem fine there. [07:10:46] <sahafeez> cool. thats what i thought [07:10:49] <d_m_c> anyone know offhand what the latest Solaris 10 kernel patch is ? Generic_118833-24 ? [07:11:52] <delewis> Generic_118855-19 on x86 [07:13:03] <Gman> richlowe, :) [07:13:12] <richlowe> Gman: see msg [07:13:24] <d_m_c> delewis: thanks .. I have a sparc server here however [07:13:49] <delewis> oh, I'm not running 10 on my Blade 1000 at the moment :-( [07:15:04] <Gadzooks> Generic_118833-30 [07:15:13] <Gadzooks> x86 [07:15:30] <d_m_c> okay .. thats a high number [07:15:31] <Gadzooks> oops [07:15:34] <Gadzooks> 118844-30 [07:15:44] <delewis> not according to SunSolve [07:17:00] <Gadzooks> delewis <- you are right for that patch [07:18:51] <d_m_c> 118833-24 SunOS 5.10: kernel patch [07:18:57] <d_m_c> thats current [07:19:54] *** pseudoXh4 has joined #opensolaris [07:20:09] <Gadzooks> dmc <- yes [07:20:41] <d_m_c> Gadzooks: in that case I am totally up to date [07:20:47] <d_m_c> thats a good thing [07:20:48] <Gadzooks> see 118822-30 as well. [07:21:25] <d_m_c> what the heck is this ? [07:21:41] <d_m_c> 18822-30 SunOS 5.10: kernel Patch [07:21:42] <d_m_c> 118833-24 SunOS 5.10: kernel patch [07:21:46] <Gadzooks> ;> [07:22:03] <Gadzooks> Crazy, huh? [07:22:08] <d_m_c> http://patches.sun.com/reports/10_patch_report [07:22:11] <DataStream> is the 22 x86 and the 33 sparc? [07:22:14] <d_m_c> I don't get it [07:22:24] <d_m_c> no .. this is Sparc only [07:22:31] <DataStream> ok [07:22:50] *** slowhog has quit IRC [07:23:41] <Gadzooks> patchdiag is your friend. [07:25:55] <d_m_c> not at the moment it aint [07:26:06] <d_m_c> Patch 120761-02 failed to install due to a failure produced by pkgadd. [07:26:17] <d_m_c> Patch 122142-01 failed to install due to a failure produced by pkgadd. [07:26:24] <d_m_c> those are Studio 11 patches [07:26:40] <delewis> d_m_c: install them with -G [07:26:59] <d_m_c> as an option to patchadd ? [07:27:04] <delewis> yes [07:27:23] <d_m_c> pkgadd: ERROR: The package <SPROprflb> is currently installed on the system in the [07:27:24] <d_m_c> global zone. To install the new instance of this package in the global [07:27:25] <d_m_c> zone only, you must specify the -G option. [07:27:28] <d_m_c> thats new [07:27:40] <delewis> it's been a problem since zones have been introduced [07:27:47] <delewis> (and patchutils became zone-aware) [07:27:58] <delewis> with pca, you can just append a -G so it installs automatically. [07:28:02] <d_m_c> ah .. lovely [07:28:05] <delewis> not surprisingly, you can't do it with smpatch. [07:28:24] <d_m_c> thankfully Sun bought aduva [07:28:37] <d_m_c> Patch 120761-02 has been successfully installed. [07:28:48] <d_m_c> Patch 122142-01 has been successfully installed. [07:29:01] <d_m_c> and people say IRC is a waste of time .. ha :-) [07:29:10] <d_m_c> thanks gentlemen ! [07:31:16] <d_m_c> hrmmm .. this is odd [07:31:37] <d_m_c> Version of package SPROmrpl from directory SPROmrpl in patch 122136-02 differs from the package installed on the system. [07:31:50] <d_m_c> very odd [07:32:33] <d_m_c> I really do wish that Sun would invent a new patch technology [07:33:09] <d_m_c> but I am sure the freakishly big government customers dictate a lot [07:36:18] <Gadzooks> d_m_c <- Sun did come out with a newer patch tool..but I don't think it is maintained anymore and wasn't well known [07:36:31] <d_m_c> oh ? [07:36:43] <d_m_c> that visual gui thing ? [07:36:45] <Gadzooks> before they had the web version [07:36:54] <d_m_c> Sun Update or whatever its called [07:37:01] <delewis> patchpro evolved into smpatch [07:37:18] <Gadzooks> that was it. [07:37:21] <delewis> everybody should be using pca, anyway :-) [07:37:39] <delewis> patchpro and smpatch are just bloated patching applications that require 50,000 Java frameworks. [07:37:49] <delewis> pca is a simple Perl script that accomplishes everything smpatch does [07:38:28] <quasi> smpatch has gotten better [07:38:38] <delewis> it still hangs [07:38:46] <quasi> not for me [07:38:57] <delewis> and it still uses 50,000 java frameworks -- there is *no* reason whatsoever to use it over smpatch. [07:39:00] <delewis> er [07:39:01] <delewis> s/smpatch/pca/ [07:39:05] <Gadzooks> delewis <- I thought the main backend of the patch tool was always PERL [07:39:22] <d_m_c> delewis: I have been using pca for months now [07:39:40] <d_m_c> but its not helping me now [07:39:55] <d_m_c> Version of package SPROmrpl from directory SPROmrpl in patch 122136-02 differs from the package installed on the system. [07:40:03] <d_m_c> I don't know what to make of that [07:40:18] <d_m_c> I just install the Studio 11 kit as per instructions [07:40:22] <delewis> you've got a different version of SPROmrpl installed than the version the patch is meant to be applied against. [07:40:27] <d_m_c> then pca fetches all the patches [07:40:34] <d_m_c> yeah .. I understand it [07:40:43] <d_m_c> I just done see it as being possible [07:40:48] <d_m_c> or probable [07:41:40] <d_m_c> it is supposed to be a patch for Sun Studio 11_x86: Performance Library Patch [07:43:52] <Gadzooks> It would be nice to have a patch tool in which you pick a particular product like Sun Studio 11 to scan for updates and get a report whether your product has the latest updates or not. helps keep the compilers and dev tools up-to-date [07:44:22] <d_m_c> pca does a great job of that [07:44:29] <d_m_c> except .. not tonight it seems [07:44:34] <Gadzooks> ;> [07:47:43] <d_m_c> its not listed as a requirement .. but I may need 120754-02 : SunOS 5.10_x86: Microtasking libraries (libmtsk) patch [07:49:41] <Gadzooks> (actually patchdiag did what I mentioned since I can remember (solaris 2.6/7) [07:50:58] <Gadzooks> d-m_c - you updating all of the Sol 10 servers? [07:50:59] <d_m_c> well .. this evening .. just for a giggle I ran pca for tunafish, the old Solaris 2.5.1 x86 server [07:51:10] <d_m_c> Gadzooks: yes Sir ! [07:55:32] <d_m_c> although isis will just get Studio patches tonight [07:56:51] <d_m_c> on that note .. I gotta get some sleep [07:57:02] <d_m_c> hopefully the NFS server will stay happy [07:57:05] <Gadzooks> me too.. [07:57:16] <d_m_c> well thanks for hanging around [07:57:22] <d_m_c> but before I go [07:57:31] <d_m_c> I need to know the answer to the riddle [07:57:37] <d_m_c> there was a riddle earlier [07:57:47] <Gadzooks> oh..therings and such [07:58:27] <d_m_c> yeah .. I figured it was a diamond [07:58:32] <d_m_c> but .. i was wrong [07:59:35] *** sahafeez has quit IRC [08:00:38] <d_m_c> I'm struck and cut, shaped and cooled, then bound by rings to release what's stored. [08:00:52] <d_m_c> I looked up the answer and I would never ever get it [08:01:09] <d_m_c> the clue was that its something that I use daily [08:01:13] <d_m_c> and I don't [08:01:20] * Auralis got it [08:01:45] <d_m_c> like .. its okay .. but not struck and cut [08:01:56] <d_m_c> they are two different terms that don't fit [08:02:00] <d_m_c> cut .. yes [08:02:01] <Gadzooks> steel [08:02:07] <d_m_c> struck .. no way [08:02:16] <d_m_c> shaped and cooled ? [08:02:17] <Gadzooks> horseshoe [08:02:18] <d_m_c> no [08:02:35] <d_m_c> shaped .. no one does that [08:02:40] <d_m_c> it gets cut yes [08:02:47] <d_m_c> but not shaped really [08:02:55] <d_m_c> oh well .. I'll get over it [08:03:07] <d_m_c> I'm struck and cut, shaped and cooled, then bound by rings to release what's stored. [08:03:15] <d_m_c> Gadzooks: you want to know ? [08:03:24] <Gadzooks> ok [08:03:25] <d_m_c> so we can both just walk away ... [08:03:28] <d_m_c> key [08:03:40] <d_m_c> bound by rings ? [08:03:41] <Gadzooks> ha! got it. [08:03:47] <d_m_c> sure .. I guess [08:04:17] <d_m_c> I shake the earth with booming thunder, fell forests whole and homes complete! I influence ships, topple kings, sweep down swift yet remain unseen! [08:04:19] *** drio has joined #opensolaris [08:04:38] <Gadzooks> interesting..always something wmple keeps us awake at night...on that note..good niht! [08:04:46] <d_m_c> ta ! [08:04:59] *** Gadzooks has quit IRC [08:05:02] <d_m_c> nite all [08:05:23] *** d_m_c has quit IRC [08:13:55] *** Symmetria has quit IRC [08:15:54] *** sahafeez has joined #opensolaris [08:16:42] <sahafeez> cork:/data - /data nfs - yes - <=- in vfstab. does not work. what am i missing. and yes cork resolves [08:17:09] <jmcp> sahafeez: has cork shared the filesystem? [08:17:23] <jmcp> do both hosts have the same domain name? [08:18:19] <sahafeez> dns domain? yes. and i let the install use the default. and i can do mount cork:/data /data from the command line and it works [08:20:03] <jmcp> did you start the nfs/client service? [08:20:10] <Byron> hi all, anyone have any experience getting Solaris to recognize onboard ethernet chips, especially the Nforce 430? [08:20:34] <jmcp> Byron: I think pretty much everybody who hangs out here has that sort of experience :) [08:20:43] <sahafeez> jmcp, there is a service i have to start to mount a share????!?! [08:20:48] <Byron> Heh, sorry first post here. [08:20:56] <delewis> sahafeez: since SBD has been enabled [08:21:05] <delewis> which is something you wanted, anyway :-) [08:21:10] <sahafeez> uhg. [08:21:15] <jmcp> Byron: no need to apologise, just letting you know you've probably come to the right place [08:21:16] <Byron> Actually, your initials look familiar, I think I read your posts on opensolaris forums about this. [08:21:23] <sahafeez> and when i do mount from a command line it does it ? [08:21:26] <jmcp> ssshhhh don't tell anybody [08:21:57] <Byron> I know an alternative is the ngo driver that Tsai (can't remember his exact name at the moment) made [08:21:59] <Byron> heh [08:22:18] <Byron> But I was hoping to get nge working [08:22:26] <Error_404> this is the most useless error ever: [08:22:29] <Error_404> SQLITE_MISUSE 21 /* Library used incorrectly */ [08:22:29] <Byron> sorry, nfo driver, not ngo. [08:22:32] <jmcp> byron: read http://blogs.sun.com/jmcp/entry/how_do_i_find_out and http://www.jmcp.homeunix.com/roller/page/jmcp?entry=how_to_find_my_nic [08:22:47] <sahafeez> jcmp - what service do i need to start and how do i make it do it at boot? [08:23:12] <delewis> nfs/client [08:23:16] <Byron> haha, that was such an old blog post that I appended my comment to, didn't think you'd see it. [08:23:48] <delewis> svcadm enable nfs/client [08:23:52] <oxygene> Error_404: as long as it only appears on really strange scenarios... [08:25:00] <sahafeez> is there a primer on the SBD stuff? [08:25:05] <jmcp> Byron: oh, it was you! [08:25:06] <delewis> primer? [08:25:14] <delewis> why do you need a primer for SBD? [08:25:41] <sahafeez> the svcadmin and new startup stuff. i am very much stuck in a 2.6 mindframe [08:25:43] <Byron> yep, one and the same. [08:25:50] <delewis> sahafeez: that has nothing to do with SBD [08:25:59] <jmcp> Byron: the other driver you were thinking of is "nfo" from Murayama-san [08:26:02] <delewis> SBD is just where the majority of SMF services are disabled [08:26:13] <delewis> specifically, the ones that start sockets [08:26:19] <delewis> er, well, bind to sockets, anyhow. [08:26:20] <Byron> jmcp: right, that's it. [08:26:40] *** loke has joined #opensolaris [08:26:52] <sahafeez> ok, well how about an smf management primer for those of us that still like startup scripts [08:26:59] <Byron> btw, when you append 'Byron:' to the beginning of your message, is that sent only to me, or does everyone see it? [08:27:02] <delewis> there's not that much to know, really [08:27:04] <delewis> svcadm disable [08:27:06] <delewis> svcadm enable [08:27:09] <delewis> svcadm milestone [08:27:12] <Byron> irc noob... [08:27:13] <jmcp> sahafeez: let me the be the first to tell you about docs.sun.com ... today :) [08:27:16] <delewis> svcs -av (to see all services) [08:27:23] <delewis> svcs -xv (to see all failed services) [08:27:25] <delewis> that's it [08:27:31] <jmcp> Byron: nah, it just means that I'm directing the comment to you. everybody sees it unless I use /msg [08:27:34] <delewis> for basic SMF usage [08:27:54] <jmcp> delewis: there's also "svcs -l [name of service]" [08:28:12] <delewis> true [08:28:17] <Byron> jmcp: ok thanks [08:28:36] <sahafeez> thanks. i want my startup scriptes back. [08:29:12] <sahafeez> time to do some reading. oh, the SBD change explains alot - what i was doing worked before and not now so i was a bit confused [08:30:06] <jmcp> sahafeez: sorry, you can't have them. they're mine! [08:30:11] * jmcp goes to make a cuppa [08:30:20] <delewis> there's a script (which I can't recall the name off of the top of my head at the moment) that's shipped that'll revert from SBD to the default, enabled services. [08:30:54] <jmcp> it's /usr/sbin/netservices open [08:30:59] <jmcp> or /usr/sbin/netservices limited [08:31:00] <delewis> there it is :-) [08:35:11] <Byron> jcmp: does the Sun Ultra 20 use the same MCP 51 chip as the Nforce 6150/430? I ask b/c you referred to it your response to my blog comment [08:35:19] <sahafeez> nice. natulais locked up [08:35:27] <delewis> dragging and dropping a file? [08:35:46] <sahafeez> nope. just clicked on a desktop icon [08:36:09] <sahafeez> hum. has not restarted - just crashed [08:38:11] <jmcp> Byron: according to http://www.opensolaris.org/jive/thread.jspa?messageID=53773 you should just be able to add your pci devids to /etc/driver_aliases, run devfsadm -i nge and then be able to see nge-attached devices in your device tree [08:38:25] <jmcp> Byron: I don't actually know. I'll see if I can figure it out [08:40:09] <jmcp> all I see from the sun system handbook is that it's the nforce4 Ultra, as delivered on the Tyan Tomcat K8E S2865 motherboard [08:41:17] <sahafeez> if i enable a service does it stay that way acrost reboots [08:41:31] <jmcp> http://www.tyan.com/products/html/tomcatk8e_spec.html says "One Broadcom BCM5721 GbE controller" and "One CK8-04 integrated MAC with Marvell 88E1111-CAA GbE PHY" [08:41:34] <jmcp> sahafeez: yes [08:41:39] <sahafeez> cool. brb [08:41:48] <jmcp> unless you use the "-t" flag to make the service state change temporary [08:42:05] <sahafeez> lockup gnome desktop nice. cannot log out.. [08:42:13] *** sahafeez has quit IRC [08:42:19] *** OnkelSchorsch has quit IRC [08:57:23] *** [1]Cyl has joined #opensolaris [08:58:44] *** [2]Cyl has joined #opensolaris [09:00:29] *** sahafeez has joined #opensolaris [09:01:03] <sahafeez> ok, dumb question - will dt wait to start until the nfs filesystem is mount. cuz i broke something if not [09:05:54] *** sahafeez has quit IRC [09:07:10] <jmcp> for anybody else who is interested, the answer to sahafeez' question is "yes" [09:12:20] *** pseudoXh4 has quit IRC [09:16:44] *** Cyl has quit IRC [09:16:44] *** [2]Cyl is now known as Cyl [09:18:25] *** [1]Cyl has quit IRC [09:20:04] *** swa has quit IRC [09:27:38] <Sieghard> cool, thanks jmcp [09:27:48] <Sieghard> gonna test it out, brb [09:28:13] *** Sieghard has quit IRC [09:30:12] *** uebayasi has joined #opensolaris [09:35:27] *** sahafeez has joined #opensolaris [09:36:38] <sahafeez> why cannot people use the same switches for nfs mounting on all OS's is beyond me. [09:38:39] <sahafeez> if i just did a mount server:/data /data from the cmdline and it worked is there away to see what options were picked? [09:40:01] <sahafeez> hum. /etc/mnttab .. cool [09:40:22] <jmcp> sahafeez: the great thing about standards is that there are so many to choose from [09:41:04] <delewis> and none of them specify how the user should interface with NFS, itself. [09:41:38] <sahafeez> bah true but nfs for god's sake. make it the same for the mount cmd switches. now since sun invented it .. [09:43:12] <sahafeez> ok, so i enabled everything that you suggested, and read thru the svcs -l for nfs and when and enabled everything marked required. on boot dt fails to start - and i get errors about trying to talk to the nfs server. [09:43:27] <sahafeez> if i comment out the line in vfstab dt starts, etc.. [09:45:03] <jmcp> sahafeez: so look in /var/dt/Xerrors for clues [09:45:10] <jmcp> and look at the logfile for the dtlogin service [09:45:15] <jmcp> find it with svcs -l dtlogin [09:45:35] <sahafeez> why would dt fail to start if i try to mount an nfs share in vfstab. that is just dumb [09:45:43] <jmcp> look in the logfile [09:47:07] <sahafeez> # svcs -l dtlogin [09:47:07] <sahafeez> svcs: Pattern 'dtlogin' doesn't match any instances [09:47:22] <sahafeez> and nothing in the Xerrors. [09:47:23] <jmcp> sorry, is this s10 or sx? s10 has dtlogin, sx has cde-login [09:47:37] <sahafeez> snv_49 [09:47:45] <jmcp> yeah, svcs -l cde-login [09:47:51] *** axisys has quit IRC [09:48:41] <sahafeez> ok, thanks. still nothing that would make sense in the log. [09:49:18] <jmcp> so when you say that dt fails to start, do you actually get the screen switching to graphics mode, or does it stay in text-mode with a message like "cannot start X server" [09:50:13] <sahafeez> hum. stops at the commandline login and dt never starts. i waited about 5 mins. got a bitch about being unable to mount the nfs share and still waiting...i was thinking that df was waiting on that... [09:51:43] <jmcp> so login on the command line and run "svcs -xv" [09:52:09] <sahafeez> ah, ok. cool cmd. [09:56:01] <jmcp> ok, time to head next door for a housewarming [09:56:02] <jmcp> ciao! [09:56:09] <sahafeez> thanks. [10:10:06] *** Gman has quit IRC [10:10:07] *** alanc_away has quit IRC [10:10:29] *** Symmetria has joined #opensolaris [10:10:51] *** alanc_away has joined #opensolaris [10:10:58] *** Gman has joined #opensolaris [10:27:13] *** sahafeez has quit IRC [10:36:38] *** klocze1 has joined #opensolaris [10:37:52] *** j0sephi has quit IRC [10:40:22] <Symmetria> lo all [10:40:37] <Symmetria> hrm question, with zfs, can you add more than one physical drive to a single pool? [10:43:58] *** swa has joined #opensolaris [10:43:59] *** spawrq has quit IRC [10:45:20] *** tsoome has joined #opensolaris [11:08:40] *** Fish has joined #opensolaris [11:10:26] <Fish> hello [11:11:49] <clee> Symmetria: of course you can [11:12:24] *** swa has quit IRC [11:15:50] *** swa has joined #opensolaris [11:17:28] *** mikaeld_ has joined #opensolaris [11:17:28] *** mikaeld has quit IRC [11:29:27] *** trs81 has quit IRC [11:44:45] *** _william_ has joined #opensolaris [11:45:15] <_william_> hi all [12:00:41] *** lplatypus has joined #opensolaris [12:03:00] <lplatypus> nexenta alpha6 won't boot up unless i remove my atheros wifi card... is this a known problem with opensolaris b50? or maybe just a nexenta thing? [12:05:10] *** Gr|ffous has joined #opensolaris [12:07:11] *** tsoome has quit IRC [12:11:05] *** Xh4 has joined #opensolaris [12:17:31] *** lplatypus has quit IRC [12:20:48] <Gr|ffous> what tools can I use to monitor cpu temp and fan speeds etc on my x86 system? [12:20:57] *** cypromis_ is now known as cypromis [12:21:08] <Xh4> Gr|ffous, you have the sensors? [12:21:28] <Gr|ffous> I used gkrellm2 used linux, to good success. [12:21:31] <Xh4> Dunno about Solaris specific but GKrellm has support for that stuff... atleast, via plugins if not by default. [12:21:33] <Xh4> Oh. [12:21:40] <Xh4> Yeah, guess I can't help you out there then. :p [12:22:24] <Gr|ffous> so you don't know if that works with solaris? [12:22:30] <Xh4> Nope. [12:22:48] <Xh4> I'm not fancy enough to own a computer which actually reports that stuff.. :P [12:23:07] <Gr|ffous> heh, well this is just an off the shelf motherboard, about 2 years old too [12:23:26] <Xh4> Oh... 2 years is pretty recent though. :) [12:23:29] <Gr|ffous> I know that solaris doesn't have cpu scaling etc yet, but the noise of the fans is really annoying me [12:23:32] <Xh4> My laptop's like 3 years old. [12:23:37] <Xh4> 4, actually. [12:23:41] <Xh4> Ah. [12:23:42] <Gr|ffous> so I'm trying to work out if solaris is just running everything at full speed all the time [12:24:28] *** SymmDC has joined #opensolaris [12:25:14] <Vanuatoo> blogs.sun.com always has response problems, does anyone know what is the problem? [12:25:33] <Gr|ffous> Following monitors are NOT supported on Solaris. [12:25:34] <Gr|ffous> CPU/Motherboard - Temperature, Voltages and Fan RPM [12:25:39] <Gr|ffous> heh - doh [12:26:10] <Xh4> Yay. :P [12:26:16] <Xh4> Limit your fan then. [12:26:27] <Xh4> Set up a resistor in series with it or something. [12:26:42] <Gr|ffous> well it's varying [12:26:52] <Gr|ffous> at the moment it's gone quiet again [12:27:19] <Gr|ffous> I've already change a couple for quieter versions, but if the CPU thinks it needs more cooling, then I shouldn't be limiting it [12:27:29] <Gr|ffous> but I really need to know the cpu temp to make that call [12:29:04] <SymmDC> ugh [12:29:10] <SymmDC> sconadm is throwing out a million exception errors [12:29:21] <SymmDC> and then dying [12:32:48] <SymmDC> hrm whats the gui equivelant of sconadm [12:40:13] <kimc> good morning [12:41:08] <SymmDC> morning [12:41:19] <SymmDC> kimc whats the gui equivelant of sconadm? [12:41:30] <SymmDC> sun sconadm keeps crashing on me when I try register [12:41:30] <SymmDC> :( [12:44:36] <sickness> morning all [12:46:22] <kimc> SymmDC: sorry dont know that one [12:47:00] <SymmDC> Exception in thread "main" java.lang.reflect.UndeclaredThrowableException [12:47:02] <SymmDC> at $Proxy1.getInstanceName(Unknown Source) [12:47:03] <SymmDC> ugh [12:51:14] *** paxc has joined #opensolaris [12:51:27] *** swa has quit IRC [12:53:09] *** swa has joined #opensolaris [12:54:47] <sickness> little question: would it be possible to run fluxbox with sunray server software? Or you are stuck with jds? [12:55:38] *** bunker has joined #opensolaris [12:56:23] <kimc> SymmDC: /bin/updatemanager [12:57:01] <SymmDC> thanks kimc [13:07:43] <andersmo> sickness: you can run whatever window manager you feel like - it's all just X to the apps. =) [13:08:07] * andersmo runs ion3 on his work sunray desktop. Works just great for packing a lot of terminals onto the screen. =) [13:08:18] <sickness> andersmo: so all the customizations are like standard dtlogin configs? [13:08:24] <sickness> cool! [13:09:46] <andersmo> Yeah, I don't think you have to do anything special on that end.. [13:11:18] <sickness> tnx! [13:16:14] <SymmDC> ergh Im wondering if I dont need to update my java or something [13:16:23] <SymmDC> cause I cant get this thing to register for shyte [13:16:54] <SymmDC> where does it actually store the registration info anyway [13:21:25] *** Risky has quit IRC [13:22:17] *** Risky has joined #opensolaris [13:23:37] <kimc> which sunray are available on ebay which would be good to learn on ? [13:23:38] *** lolownia has joined #opensolaris [13:28:55] <lasseoe> kimc, any of them [13:35:37] *** swa has quit IRC [13:44:13] *** vmhobbes has joined #OpenSolaris [14:08:42] *** vmhobbes has quit IRC [14:16:09] *** Arnald has joined #opensolaris [14:17:18] *** vmhobbes has joined #OpenSolaris [14:21:22] *** swa has joined #opensolaris [14:23:58] <astinus> I've got a problem with audio output [14:24:09] <astinus> I was playing a XviD file yesterday, and it just stopped working [14:24:16] <kimc> lasseoe: ok very good so none of them are 'undesireable' [14:24:17] <astinus> I can't even get mplayer to output simple music :/ [14:24:44] <kimc> brb.. [14:25:33] <astinus> Anyone tell me how I can kick Solaris into doing audio again ? :/ [14:29:30] <astinus> cat /dev/urandom >> /dev/sound/0 works good [14:30:59] <astinus> okay, weird .. works now :X [14:33:24] *** salamanders has joined #opensolaris [14:37:40] <sickness> try something like this? [14:37:41] <sickness> audioctl -v 100 [14:37:41] <sickness> audioctl -p speaker+headphone+line [14:37:41] <sickness> audioctl -U [14:37:54] <sickness> I used to do that on a crappy ac97 on a crappy dell ;) [14:48:05] *** salamanders has quit IRC [14:48:06] *** klocze1 has quit IRC [14:54:24] <movement> any US SWAN users about? [14:55:35] <movement> never mind [14:55:36] <ProfMikey> movement: what's the matter? [14:57:07] *** mikaeld has joined #opensolaris [14:58:41] <movement> ProfMikey: doesn't matter, I found out on gcn [14:59:46] <ProfMikey> ah, k :) [15:10:51] *** lejocelyn has joined #opensolaris [15:10:58] <lejocelyn> hi [15:11:10] *** mikaeld_ has quit IRC [15:11:19] <jamesd> hi [15:11:41] <lejocelyn> is there a french channel for opensolaris ? [15:12:23] <jamesd> not that i know of.. but we do have a few french speaking members [15:12:43] <lejocelyn> nevermind [15:12:48] <lejocelyn> it's not a big problem [15:13:47] <lejocelyn> I'm just seeking informations about opensolaris on the net [15:14:05] <lejocelyn> and may-be I'll ask something [15:14:08] <lejocelyn> after [15:14:28] <jamesd> unixconsult.org/solaris is a good search engine for finding stuff related to opensolaris/solaris or you can check out our homepage at opensolaris.org [15:14:54] <lejocelyn> yep, I'm already their [15:14:55] <lejocelyn> there [15:14:56] <Auralis> just ask if ya have questions [15:15:55] <lejocelyn> :) ok [15:16:17] <lejocelyn> I'm just wondering what are the difference between the sun's open licence and the GPL [15:16:56] <lejocelyn> I see that there are some projects like GNU/Sunos [15:17:09] <lasseoe> lejocelyn: http://blogs.sun.com/chandan/category/Solaris [15:17:11] <jamesd> http://www.opensolaris.org/os/about/faq/licensing_faq [15:17:19] <lejocelyn> thanks [15:18:04] <lejocelyn> I heard that the sun's filesystem is really good [15:18:16] <lejocelyn> but I don't remember its name [15:18:21] <jamesd> zfs is great [15:18:24] <lejocelyn> :/ [15:18:28] <jamesd> but can't be booted yet. [15:19:06] <lejocelyn> ah ok [15:19:27] *** Somethingelse has joined #opensolaris [15:20:19] <SymmDC> bootable zfs would be awesome [15:20:26] <myrkraverk> erm, what is solaris for "setserial" ? [15:20:56] <SymmDC> heh intel needs to start making servers that boot in < 10 minutes :p [15:21:01] <jamesd> SymmDC, its coming... [15:21:23] <SymmDC> heh james Im hoping that in the next major release of solaris as well they will include the lsi megaraid drivers [15:21:29] <jamesd> SymmDC, why? you only boot a machine maybe 1/2 a dozen times a year [15:21:47] <jmcp> unless you're using it for dev work [15:21:53] <jmcp> like writing drivers [15:22:04] <SymmDC> heh james when you're troubleshooting and the rest of it [15:22:10] <SymmDC> you can reboot a lot more often than that [15:22:20] <SymmDC> particularly when you're working on the things raid arrays [15:22:20] <SymmDC> :p [15:22:46] <SymmDC> hrm [15:22:49] <SymmDC> this is very odd [15:23:13] <jamesd> no one talks of making starfire 25k's... [15:23:16] *** mazon is now known as Mazon [15:23:17] <jamesd> boot faster [15:23:32] <jmcp> jamesd: actually, I think they do [15:23:48] <jmcp> mainly the response comes back "turn off max-level post you numbskull!" [15:23:49] <jmcp> :) [15:24:01] <jamesd> :-) [15:24:06] <SymmDC> hrm this is odd, when I create a vlan interface, if I do it by script, create it, and up it straight away [15:24:09] <SymmDC> it doesnt come up [15:24:19] <SymmDC> I have to run something like this: [15:24:20] <SymmDC> for i in `ifconfig -a |grep e1000g2000 |awk '{print $1}' |sed 's/:$ [15:24:20] <SymmDC> //g'`; do ifconfig $i up; done [15:24:26] <SymmDC> in order to actually bring all the sub interfaces up [15:30:16] <paxc> SymmDC : better if you use awk qith -F":" option [15:30:24] <lejocelyn> yep, OpenSolaris seems interesting, I' guess one day I'll test it [15:30:28] <paxc> you will not need sed redirect [15:30:35] <lejocelyn> when I'll have to set up a server [15:31:05] *** lejocelyn has left #opensolaris [15:31:47] <SymmDC> heh pax depends, in this particular case the sed works well [15:31:47] <paxc> ho. you will need -__ [15:31:49] <paxc> sorry [15:31:59] <SymmDC> but its just odd that I even have to run that [15:32:02] <SymmDC> to bring the interfaces up [15:33:22] <paxc> you will not need in this casa sed :) [15:33:25] <paxc> case [15:33:32] <paxc> but works in the same way :) [15:35:27] <SymmDC> pax, yeah but still think sed is cleaner than the equivelant |awk -F: '{print $1":"$2}' [15:35:49] <SymmDC> because the sed Im using there simply removes one :, I still need column 1, the : between it, and column 2 :) [15:36:25] <myrkraverk> what do I use in solaris to set the baud of a serial port? [15:36:49] *** spawrq has joined #opensolaris [15:41:58] <SymmDC> hrm [15:42:05] <SymmDC> can someone see if they can ftp to biomirror.mirror.ac.za [15:42:10] <SymmDC> and if they can download ok etc [15:42:19] <SymmDC> just wanna make sure thats running again 100% from outside [15:42:54] *** vmhobbes has quit IRC [15:47:20] <paxc> myrkraverk : ttyadm [15:47:31] <paxc> belive [15:50:45] *** vmhobbes has joined #OpenSolaris [15:51:24] <myrkraverk> paxc: I don't seem to find out how from the man page [15:51:37] * SymmDC blows up google [15:53:46] *** koolniczka has joined #opensolaris [15:54:26] <paxc> myrkraverk [15:54:29] <paxc> what you need? [15:55:32] *** klocze1 has joined #opensolaris [15:55:51] <myrkraverk> paxc: I wanted to set my serial port to 115200 baub, before running pppd on it [15:56:08] <myrkraverk> in linux, I'd use setserial for this [15:56:27] <paxc> myrkraverk : you can also change from /etc/ttydefs by hand or use smc ( solaris 10 ) to change [15:56:36] <myrkraverk> ok [16:00:03] *** vmhobbes has quit IRC [16:01:02] *** paxcs has joined #opensolaris [16:03:32] *** SymmDC has quit IRC [16:07:15] *** alobbs has joined #opensolaris [16:09:10] <astinus> Anyone care to suggest why X.org is over 300MB of RAM? :/ [16:10:42] <myrkraverk> astinus: becoust it counts your video ram too [16:10:55] <myrkraverk> (afaik) [16:11:07] *** vmhobbes has joined #OpenSolaris [16:13:13] *** jcsmith has joined #opensolaris [16:13:41] <astinus> 0827F000 290300 223992 223936 - rw--- [ heap ] [16:13:52] <astinus> claims its using 220Mb resident, for its heap [16:18:12] *** paxc has quit IRC [16:18:17] *** paxcs is now known as paxc [16:21:50] *** jcsmith has quit IRC [16:22:21] *** salamanders has joined #opensolaris [16:25:59] *** jumpstart has joined #opensolaris [16:31:50] *** paxc has quit IRC [16:41:08] *** bengtf__ has joined #opensolaris [16:41:12] *** bengtf__ is now known as bengtf [16:45:37] *** tsoome has joined #opensolaris [16:49:53] *** nealc has joined #opensolaris [16:50:28] <myrkraverk> erm, what do I use to get a list in solaris, like I do wih plain "ifconfig" in linux? [16:50:52] <jamesd> ifconfig -a or netstat -in [16:51:43] <myrkraverk> ok, thanks ;) [16:59:32] <tsoome> myrkraverk: read the fucking manual? [17:02:07] <PerterB> why read when you can bug people on IRC ;) [17:03:57] *** Symmetria has quit IRC [17:05:05] <tsoome> sometimes I think they just can't read....;) [17:08:38] *** miffe has joined #opensolaris [17:09:45] *** delewis has quit IRC [17:10:02] *** lisppaste3 has quit IRC [17:12:47] *** lisppaste3 has joined #opensolaris [17:21:32] *** nealc is now known as nlc|afk [17:35:15] *** bengtf has quit IRC [17:37:11] *** Darwin_ has joined #opensolaris [17:45:45] *** Arnald has quit IRC [17:46:52] *** jsubl2 has joined #opensolaris [17:49:46] *** nachox has joined #opensolaris [17:50:10] *** Darwin has quit IRC [17:51:13] *** jumpstart has quit IRC [17:52:08] *** mega has joined #opensolaris [17:53:06] *** jteo has quit IRC [17:58:20] *** Darwin has joined #opensolaris [18:02:54] *** klocze1 has quit IRC [18:05:07] *** swa has quit IRC [18:07:59] *** mega has quit IRC [18:10:18] *** Darwin_ has quit IRC [18:13:40] *** jteo has joined #opensolaris [18:13:49] *** kloczek has joined #opensolaris [18:17:15] <gnu2it2> i have a headless ultra 5. is there an east way to free up some space by removing the CDE files? [18:17:50] <jamesd> find another old ide drive and install it... [18:18:02] <jamesd> you can get an 80 GB drive for a few bucks. [18:18:32] <gnu2it2> have a 60 maxtor,, need to figure out ho to get a label on it :) [18:18:57] <jamesd> format -e [18:19:58] <gnu2it2> core dumps whane selected from format -e [18:20:42] <jamesd> install sformat it may do better [18:33:23] <jteo> darn. zfs jus hardlocked my box. gotta love ARC. [18:34:31] <_william_> is there an easy way to format (low level) a disk with sformat (easy is important in this sentence) [18:35:05] <jamesd> i've never had to .. only had to label a disk.. perhaps the man page. [18:36:17] <_william_> not found in the man page. I was looking for a way to format and change the sector size of an hard disk, and it was such a mess :( i had to look for hundreds of parameters from manufacturer documentation [18:36:24] <_william_> finally i formatted it using IRIX ;) [18:37:14] <jteo> great. now the zpool doesn't come up. [18:37:43] <nachox> i hope you didnt have important data there... [18:38:18] <jteo> nachox, i was running a benchmark. [18:39:39] * nachox wonders if solaris 10u3 will include support for intel's 950 graphic cards [18:40:16] <jteo> nachox, unlikely. [18:42:17] <nachox> they work with the vesa driver anyway right? [18:44:46] <jteo> nachox, i presume so. [18:45:43] <nachox> i guess that means slow X for me, well, i can live with that [18:48:11] *** nlc|afk is now known as nealc_ [18:49:18] *** delewis has joined #opensolaris [18:50:19] <jteo> things don't look good. maybe i'm having...hard disk failure. [18:51:21] <Tekni> jteo: what kind of hardware do you have? [18:51:52] <jteo> Tekni, just 2 PATA disks in a Shuttle shoebox. [18:52:02] <jteo> UFS log replay just failed. [18:52:28] <Tekni> zfs doesn't seem to like cheap stuff as much as its touted to [18:53:37] <Tekni> and it seems hard to tell whether its disk, controller, or zfs bug on the failures i'm getting [18:53:55] <jteo> Tekni, right now, i'm just waiting for fsck to finish. -sigh- [18:55:41] <jteo> though i'm trying to figure out why a hard reset on a hung zpool would affect my UFS drive. [18:55:43] <jteo> :( [19:03:18] *** uebayasi has quit IRC [19:03:40] <_william_> anyone knows if solaris 10 works on hp LP1000 server ? [19:05:13] *** nachox has quit IRC [19:12:01] <jteo> i'm reminded once again how frail hardware can be...-sigh- [19:12:50] *** mega has joined #opensolaris [19:13:26] <g4lt-U60> _william_, is it on ht eHCL? [19:14:16] *** sahafeez has joined #opensolaris [19:25:05] *** LeftWing has quit IRC [19:25:16] *** LeftWing has joined #OpenSolaris [19:28:25] <Error_404> bloody sqlite3_close() segfaults my code [19:30:41] *** lolownia has quit IRC [19:31:21] *** aliquis has quit IRC [19:42:33] *** kloczek has quit IRC [19:42:58] *** deather__ has joined #opensolaris [19:50:06] <_william_> yes it is thanks g4lt-U60 i havent seen it the time i looked :) [19:50:10] <_william_> too fast certainly ;) [19:50:13] <_william_> cool [19:59:29] *** loke has quit IRC [20:00:25] *** deather_ has quit IRC [20:03:17] <sickness> evening all [20:03:38] <_william_> evening sickness [20:04:24] *** simford has quit IRC [20:05:03] *** Symmetria has joined #opensolaris [20:05:43] <Symmetria> lo all [20:08:16] <jteo> wb Symmetria. [20:22:46] <Symmetria> has anyone here built a php5 module for apache 2.2 [20:22:59] <Symmetria> or have a php5 module that will work with teh blastware installed version of apache [20:25:18] *** mega has quit IRC [20:27:15] *** mega has joined #opensolaris [20:28:41] *** LeftWing_ has joined #OpenSolaris [20:28:52] *** LeftWing has quit IRC [20:30:07] *** crib2 has joined #opensolaris [20:31:53] *** Fish- has joined #opensolaris [20:34:53] *** kloczek has joined #opensolaris [20:47:11] <_william_> Symmetria, the php5 module from blastwave does not work ? [20:47:55] <gnu2it2> which slice on a disk should be the whole disk? [20:48:05] <_william_> 2 [20:48:31] *** Fish has quit IRC [21:08:07] *** mega_ has joined #opensolaris [21:10:23] *** mega has quit IRC [21:10:53] <jteo> hmm. gentoo is certified for T1000 and T2000 [21:11:24] <oxygene> Symmetria: I have php5 for solaris 10's version of apache2 [21:12:14] *** bunker has quit IRC [21:15:05] *** kimc has quit IRC [21:15:56] <jamesd> jteo, certified by who? and who is going to support it? is it certified for any middleware or database? [21:16:29] <jteo> jamesd, Sun. [21:17:22] <jteo> support provided by the Gentoo Community. ooh. [21:17:31] <jamesd> sun supporting gentoo? who is smoking crack? gentoo breaks every other minute. [21:18:15] <sickness> gentoo with linux kernel or opensolaris base? (like nexenta) [21:18:49] <oxygene> gentoo with linux [21:18:53] <jteo> sickness, Gentoo Linux [21:18:58] <sickness> uhm [21:19:11] <jteo> jamesd, I meant Sun certified Gentoo Linux as running on the T1000 and T2000. [21:27:05] <_william_> you mean this T1000 was running Gentoo Linux ? :) http://en.wikipedia.org/wiki/T-1000 [21:28:44] *** claudiush has joined #opensolaris [21:29:03] <sickness> http://gentoo-wiki.com/Alba_Experiment <- I'd be curious about this project [21:29:15] <sickness> if there's a thing that I really dislike of gentoo, it's the linux kernel :') [21:30:37] <sahafeez> linux issuse is the lack of it being a true OS. it is a kernel with some userland strapped on. that being said, gentoo is the best version of linux. i used to say slackware but portage rocks. [21:31:09] <delewis> and what OS isn't a kernel with a userland strapped on it? [21:32:21] <sahafeez> any os you can do make build world on [21:32:42] <sahafeez> any os that the userland breaks if the kernel is not synced to it, openbsd, freebsd, etc.. [21:32:50] <delewis> you can do that on Linux distributions that are structured in a manner that allows it. [21:32:57] <delewis> you being a Gentoo user should know this. [21:33:15] <delewis> I'm just saying your definition of "true OS" and the criteria for it are highly-subjective. [21:33:23] <myrkraverk> erm, where/how do I config nameservers on solaris? [21:33:26] <sahafeez> yes, and you point is ;) [21:33:36] <sahafeez> nsswitch.conf [21:33:39] <sahafeez> resolv.conf [21:34:21] <myrkraverk> and if I don't have an /etc/resolv.conf, do I make it? [21:34:26] <sahafeez> yes [21:34:36] <tsoome> myrkraverk: there is a very nice site. it's called docs.sun.com. [21:34:51] <myrkraverk> tsoome: it is also very large ;) [21:34:54] <sahafeez> need to set hosts: files dns [21:34:54] <sahafeez> in nsswitch [21:34:59] *** bunker has joined #opensolaris [21:35:02] <sahafeez> then make /etc/resolv.conf [21:35:42] <tsoome> yes, it is, but there is also little thing - it's called search. [21:35:53] <myrkraverk> tsoome: ;) [21:37:49] <jteo> so much loving going on in here. [21:38:47] *** mega__ has joined #opensolaris [21:39:00] *** mega_ has quit IRC [21:40:10] <myrkraverk> hmm, I can ping my 192.168.0.5 - but even with it in /etc/resov.conf - it does not seem to work ;/ [21:40:22] <quasi> gotta get more of that, or something [21:40:46] <tsoome> myrkraverk: how you know it does not wotk? [21:40:58] <tsoome> I mean, what commands? [21:41:08] <myrkraverk> tsoome: dig www.google.com @192.168.0.5 [21:41:17] <myrkraverk> nslookup www.google.com [21:41:47] <tsoome> well then check if there is a name server up and running;) [21:41:56] <myrkraverk> it works from my linux box [21:42:22] <tsoome> dns was not configured in solaris before? [21:42:31] <tsoome> then restart nscd and try again [21:42:37] <Berny> look if you got dns als method in /etc/nsswitch.conf for looking up hosts [21:43:12] <tsoome> Berny: nslookup/host/dig are using libresolv directly IMO [21:43:49] *** silly_girl22 has joined #opensolaris [21:43:50] <myrkraverk> how do I restart nscd? [21:44:29] <myrkraverk> (yes, I'm a solaris n00b) [21:44:36] <Doc> what version of solaris? [21:45:07] <myrkraverk> excr b50 [21:45:34] <Doc> svcadm restart name-service-daemon or somethinbg stupid like that [21:46:05] <Doc> svcs '*name*' should help [21:46:14] *** kloczek has quit IRC [21:46:51] <quasi> prolly svc:/system/name-service-cache:default [21:47:13] <myrkraverk> yes, that's what I tried [21:47:16] <dwc-> 'pkill nscd' and let smf take care of restarting it ;) [21:47:44] <Doc> dwc: ugly, but yes it will work as long as you dont do it too many times [21:47:47] <quasi> dwc-: evil! [21:48:13] <dwc-> <-- lazy :) [21:48:16] <sahafeez> any reason to use sunstudio 10 over 11...there used to be something.. [21:48:35] <delewis> many reasons -- better code generation on AMD64 being the primary. [21:48:49] <delewis> Sun Studio 11 is also license-free. [21:49:29] <delewis> and IIRC, Sun Studio 11 supports AMD64's various memory models (10 might've supported this -- I'm not sure) [21:49:37] <sahafeez> ok, cool. [21:49:42] <delewis> regardless, there's a "What's New" section in the Sun Studio 11 documentation [21:49:58] <myrkraverk> hmm, none of that seemed to help (and yes, I have "hosts: files dns" in my nsswich.conf) [21:49:59] <sahafeez> if i am running 49 and want to go to 50 with out an reinstall and live update will not work ... [21:50:01] <quasi> sahafeez: no, no reason at all - 11 is "more free" than 10 [21:50:04] <dwc-> wait, those sound like reasons to use 11 over 10? [21:50:30] <delewis> oh, I meant reasons to use 11 over 10 :-) [21:50:31] <myrkraverk> is there any difference from amd64 and intel x64? [21:50:41] <sahafeez> you could not build ON in SS11 or something like that back when - racking brain. knows he read something.. [21:50:43] <clee> myrkraverk: very small differences [21:50:44] <delewis> myrkraverk: slight differences, but very slight ones. [21:50:47] <clee> myrkraverk: mostly in favor of AMD [21:50:58] <delewis> EM64T (for the most part) complies with AMD64 [21:51:13] <delewis> most of the differences are because of gaps in the specification, IIRC. [21:51:22] <myrkraverk> so - should not have to think about it, and just compile my stuff as amd64? [21:51:25] <sahafeez> enough that if you take an amd64 distro install and put the disk in an intel64 it will not boot [21:51:28] <quasi> dwc-: from something like 46 or 47 you need 11 to build osol and on the license side, 10 is free for osol community, 11 is free for anyone signed up for sun.com/developers [21:51:34] <delewis> myrkraverk: no [21:51:43] <delewis> gcc for instance will do pipeline optimizations differently [21:51:55] <myrkraverk> sahafeez: what do you mean? I *am* running excr as amd64 [21:52:18] <myrkraverk> does SS11 difference it too? [21:52:22] <delewis> the spec doesn't specify how long the pipeline should be or what the pipeline should do at eacn stage -- that would be silly. [21:52:25] <delewis> myrkraverk: yes [21:52:32] <myrkraverk> ok [21:53:10] <sahafeez> if you take openbsd or freebsd amd64 distro installed on a harddrive and put it in an intel64 box it will tank. if you take the i386 version and move boxes it works just fine. [21:53:29] <delewis> myrkraverk: it won't make that much of a difference, but if you want the most performance out of your application, it will. [21:53:52] <delewis> sahafeez: sad. [21:54:06] <myrkraverk> delewis: ok ;) [21:54:33] <sahafeez> makes sense. both use the Wx bit on amd64 [21:55:08] <delewis> Solaris works fine on both AMD64 and EM64T :-) [21:55:30] <sahafeez> it is not built with amd64 only stuff turned on. [21:55:50] <delewis> ? [21:56:19] <delewis> Solaris supports non-executable stacks on both [21:56:59] <sahafeez> ok. well i am not a prg'er. just what i read. [21:57:13] <sahafeez> nfsd send error 55 ? [21:58:00] <delewis> 08047000 4K rw--- [ stack ] [21:58:07] <delewis> 0808A000 32K rw--- [ heap ] [21:58:14] <delewis> that's on an EM64T system (output from pmap) [21:58:38] <delewis> the stack and heap both are missing their execution bits [21:59:08] <sahafeez> cool [22:02:20] <Symmetria> quick question, how do I add various paths to my library search path for normal apps (so they will check places like /opt/csw/lib etc when they try start) [22:02:45] <delewis> crle, but you shouldn't need to take such measures if your applications are built properly. [22:02:56] <delewis> and regardless are you referring to compile-time linking or run-time linking? [22:03:27] <Symmetria> run time stuff [22:03:44] <Symmetria> when it runs and it looks for a library file [22:03:49] <Symmetria> ld.so.1: httpd: fatal: libgcc_s.so.1: open failed: No such file or directory [22:03:50] *** mega__ has quit IRC [22:03:56] <delewis> you can use crle to add the paths, but like I said, try building your applications first properly by including the run-time linker search path in the executable. [22:04:11] <Symmetria> delewis these are packages from sunfreeware etc that Im installing [22:04:29] <delewis> you need to install Sun Freeware's gcc or gcc-lib package. [22:05:21] <delewis> the Sun Freeware packages are built with embedded run-time linker paths and it's only searching those, trying to find libgcc_s.so.1 which is missing. [22:05:24] <Symmetria> heh delewis I've got a lot of the packages installed from blastware, but the library location is different :) [22:05:37] <jengelh> CSW does the libgcc thing right [22:06:02] <Symmetria> hrm ok lemme install the sunfreeware versions [22:12:02] <myrkraverk> can solaris have a frame buffer image on boot - similar to how some linux distros do (on x86) ? [22:12:58] *** onlinebacon has joined #opensolaris [22:13:04] <jamesd> if you re-write the kernel loader you can do anything you want. [22:13:27] *** gisburn has joined #opensolaris [22:13:31] <gisburn> Hi! [22:13:39] <myrkraverk> gisburn: ;) [22:13:40] <jamesd> hello [22:13:41] <gisburn> Another quiz: "I shake the earth with booming thunder, fell forests whole and homes complete! I influence ships, topple kings, sweep down swift yet remain unseen!" - what am I ? [22:13:42] <myrkraverk> jamesd: hmm, ok ;P [22:14:30] <onlinebacon> gisburn: Wind? [22:14:56] <gisburn> umpf [22:14:58] <gisburn> onlinebacon: yes [22:14:59] <gisburn> ;-( [22:15:49] <dwc-> keys! [22:15:53] <dwc-> wait... wrong riddle [22:16:19] <dwc-> is this one from stargate too? [22:16:27] * gisburn hits dwc- with a stick [22:16:31] <onlinebacon> woo [22:16:49] <gisburn> dwc-: yes, but the "key" one was from the original authurian legend. [22:16:56] <Symmetria> hrm anyone have any idea why I have to run an up command against every ip alias I add on a vlan, and when I do add them, if I issue the up command 2 fast it doesnt come up [22:17:03] * dwc- beats gisburn's komodo dragon with with the stick [22:17:11] <myrkraverk> gisburn: wind? I was going to say air (but not a whole lot of difference there) [22:17:14] * onlinebacon hides because there is a dragon [22:17:25] * myrkraverk goes off to do stuff [22:17:34] * dwc- tells gisburn's komodo dragon to go eat the alien bug invading germay [22:17:38] * onlinebacon looks for a girl called stuff to do too [22:17:53] <dwc-> stuffed? [22:18:02] <onlinebacon> a girl called stuff, to do [22:18:05] <onlinebacon> ah dw bout it :) [22:18:42] <sickness> hi gisburn :) [22:18:48] * delewis watches gisburn's head explode [22:19:06] <_william_> a riddle... a riddle ? :) "Who would cross the Bridge of Death must answer me these questions three, ere the other side he see." [22:19:54] <sahafeez> holly grail, something and blue [22:20:02] <_william_> :) [22:20:28] <sahafeez> aferican or eu swallow? [22:20:31] <quasi> sahafeez: who is Holly and what grail are we talking about? ;) [22:20:36] <sahafeez> wow! i can speel today [22:20:51] <_william_> sahafeez, "What... is your favourite colour?" :) [22:21:23] <quasi> _william_: that's a trick question ;) [22:21:24] <sahafeez> blue, no green, .... aahghghgh! [22:21:33] *** jcsmith has joined #opensolaris [22:22:04] <_william_> so quasi ... "What... is the capital of Assyria?" :D [22:22:38] <onlinebacon> anyways off to install os, cya later [22:22:39] *** onlinebacon has quit IRC [22:22:40] <dwc-> A? [22:28:33] *** crib2 has quit IRC [22:29:54] <Berny> atin what order are filesystems mounted? imagine i have a zfs containing a directory which shall be mountpoint for an ufs... would that give problems when i boot that box? [22:32:06] <Symmetria> woooooooooot [22:32:13] <Symmetria> mirror.ac.za is back up finally on solaris [22:32:22] <Symmetria> now its just the sync scripts and Im all done :) [22:32:23] <sahafeez> you can set the boot order in vfstab if i remember correctly [22:32:33] <sahafeez> mount order [22:33:20] *** gisburn has left #opensolaris [22:36:26] *** jteo has quit IRC [22:44:23] <Berny> ok next question: besides filesystem quota on zfs can i have per user quota inside a single zfs without having single filesystems for each user? (imagine /var/mail) [22:47:40] <richlowe> Nope. [22:48:01] <Berny> bugger [22:48:12] <richlowe> yeah, it's been a subject of some fairly heated debate on zfs-discuss. [22:48:27] <Berny> i vote to put that on the wishlist ;-) [22:48:40] <richlowe> it makes accounting problematic. [22:49:25] <richlowe> (how does data in snapshots get charged? does it get charged at all?, etc.) [22:49:29] <dwc-> what? you can make /var/mail/username a symlink to /var/mail/filesystems/user/user ;) [22:49:41] <dwc-> for all 20,000 users ;) [22:49:57] <richlowe> The most filesystems I've ever had at once has been 2,500-ish [22:50:04] *** logic_ has quit IRC [22:50:09] <Berny> i'd rather shoot myself in both knees ;-) [22:50:10] <richlowe> I think I could have gone higher, but it has a noticable effect on boot time. [22:50:29] <richlowe> Other than that, there isn't much down side, at least in the 2,000-3,000 range [22:52:17] <richlowe> and someone with enough anonymous tracing-fu, and the desire to make that better probably could. [22:52:34] <richlowe> I'd wager nobody has really thought about mounting many-thousands of filesystems as a boot-time performance thing, pre-zfs :) [22:53:54] <Berny> so far i have about 10 ufs, 3 samfs, 5 zfs ;-) [22:55:28] *** logic has joined #opensolaris [22:56:40] <Symmetria> heh zfs is working liek a dream for me [22:56:45] <Symmetria> and I have some *HUGE* file systems on it [22:56:52] <Symmetria> apool 3.8T 2.6T 1.2T 68% /mirror/iscsi-san-1 [22:56:52] <Symmetria> bpool 3.8T 24K 3.8T 1% /mirror/iscsi-san-2 [22:56:52] <Symmetria> cpool 3.8T 24K 3.8T 1% /mirror/iscsi-san-3 [22:56:52] <Symmetria> dpool 1.1T 50G 1.0T 5% /mirror/internal [22:57:04] <Berny> it's doing a nice job ofr my home on the notebook [22:57:18] <Berny> no it's time to get this into the real world ;-) [22:57:19] <g4lt-U60> anything crossing the zetabyte boundary yet? [22:57:23] <g4lt-U60> *yawn* [22:57:33] <richlowe> Yeah, I wasn't talking about size of filesystem, but number thereof. :) [22:57:41] <richlowe> create thousands, and watch what it does to boot time :) [22:57:47] <Symmetria> heh, what impresses me most about zfs [22:57:54] <Symmetria> is when you start doing very i/o intensive stuff [22:58:00] <Symmetria> it handles well [22:58:03] <Symmetria> with millions of tiny files [22:58:25] <richlowe> (if I ever get a real console back on this machine, not the damn IPKVM, I'm going to create a whole lot more, and try and get some sane idea *where* it's slowing down) [22:58:34] <richlowe> or rather, what it'd take to making mounting faster. [22:58:34] <Symmetria> that iscsi-san-1 has god knows how many tiny files on it [22:58:48] <dwc-> serial console [22:59:00] <richlowe> dwc-: Yeah, but then the term server died. [22:59:04] <richlowe> dwc-: I had the choice between IPKVM or nothing. [22:59:07] <Symmetria> heh richlowe be lucky you can use the machines console, my machine wont boot off anything other than serial console :p [22:59:20] <Symmetria> it freaks on my vga card [22:59:25] <Berny> Symmetria: i had like 26mio files on a samfs ;-) [23:00:06] <Symmetria> hrm, how do I see number of files on a file system, df -i doesnt work [23:00:13] <Berny> dunno [23:00:20] <Berny> samfs is nice with samu :-) [23:00:37] <Berny> regular files 9,321,532 4.813T 5292459019118 [23:01:14] <Berny> number of files/ size in M/G/T / size in bytes that is [23:01:17] <Symmetria> heh Im about to see just how well solaris handles io though, about to file up a complete mirror sync [23:01:22] <Symmetria> that hasnt been synched in 4 days [23:01:33] <Symmetria> and most of that is rsync and rsync can be horribly heavy i/o wise [23:02:02] <jengelh> no way [23:02:25] <jengelh> well it stat's all files first, though. [23:05:12] *** nealc_ has quit IRC [23:05:55] *** Somethingelse has quit IRC [23:11:48] <Symmetria> heh [23:11:56] <Symmetria> seems I got my tcp tuning variables right under solaris [23:11:57] <Symmetria> 367001600 bytes transferred in 38 seconds (9.20M/s) [23:11:57] <Berny> bugger, how can i cancel i resync on svm mirrors when i want to delete that mirror? [23:12:05] <Symmetria> considering thats over 400ms latency [23:12:08] <Symmetria> cross continent [23:12:54] *** gnu2it2 has quit IRC [23:14:02] *** jsubl2 has quit IRC [23:18:34] <Berny> .oO(why doesn't metasync on sol10 have the -c option yet?) [23:21:51] *** Administrator has joined #opensolaris [23:29:06] *** nwf has joined #opensolaris [23:31:07] *** richlowe has quit IRC [23:32:43] *** richlowe has joined #opensolaris [23:42:07] *** knightblader has joined #opensolaris [23:48:24] *** laca has joined #opensolaris [23:59:31] *** jafari has joined #opensolaris