July 10, 2011 
< | 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 | >
July 10, 2011 [00:00:03] <tsoome> if you really have to mount it to /home, sure, you can. but as /home is automounter mount point by default, you need to reconfigure it. and accept that may cause some issues if original entry in auto_master will get restored:P[00:00:40] <DanaG> hmm, perhaps there's a more elegant way to do it, such as editing the user's profile. So we'd use /export/home[00:00:44] <DanaG> s/we/I/[00:00:55] <tsoome> ?[00:01:28] <tsoome> well, you may wanna look into /etc/auto_home as well, and still read man;)[00:03:29] <DanaG> Will do, once I get the server booted again. Doing something with it at the moment, so it'll be a while.[00:03:31] <DanaG> Thanks for the tips.[00:04:24] <tsoome> thats for the case you still wanna user homes to be in /home but physical location in /export/home ;)[00:04:47] <DanaG> ah, so zfs would make /export/home, and autofs would map that to /home.[00:05:05] <tsoome> normally user by user.[00:05:51] <tsoome> as you dont wanan have /home/home/user but /home/user[00:06:46] <tsoome> automounter does mount when the "key" is accessed.[00:07:22] <blues> any of you have the sasuc8i?[00:11:39] *** Oriona has quit IRC[00:13:34] *** Oriona has joined #openindiana[00:13:49] *** xmikus01 has joined #openindiana[00:34:32] *** DrLou has joined #openindiana[00:34:32] *** ChanServ sets mode: +o DrLou[00:35:44] *** dijenerate has quit IRC[00:49:00] *** Naresh``` has quit IRC[01:05:21] *** hsp has quit IRC[01:10:25] <DontKnwMuch> blues: which card do you have? I have flashed some m1015 sucessfully[01:11:05] <DontKnwMuch> is this sasuc8i a lsi1068 or lsi2008[01:11:43] <DontKnwMuch> ah, it is 1068[01:13:40] <DontKnwMuch> blues: your problem is your mainboard, try to flash the card in another mainboard[01:24:49] *** dijenerate has joined #openindiana[01:31:13] <RoyK> 1068 has a few issues with OI/solaris[01:31:43] <RoyK> or the other way around[01:33:19] <DontKnwMuch> I thought these were used by sun on some servers they made, and were supported fully[01:34:49] <blues> Dontknwmuch, my mainboard? really?[01:35:03] <DontKnwMuch> yes.[01:35:08] <DontKnwMuch> I am quite sure[01:35:24] <blues> you wouldn't think you could get much more vanilla than an intel board. What causes the problem, do you know?[01:35:31] <richlowe> Wait, what's the problem now?[01:35:36] <DontKnwMuch> Yes exactly an intel board[01:35:47] <DontKnwMuch> 3000/5000 chipset perhaps?[01:35:54] <blues> p55[01:35:57] <blues> i think[01:36:03] <richlowe> RoyK: Your configuration is weird as hell, I still don't think blaming the board is valid[01:36:15] <richlowe> compared to blaming your combination of expensive card, expananders, and cheapass disks.[01:36:26] <blues> yeah, p55... specifically DP55WG[01:37:05] <RoyK> richlowe: my configurations aren't very weird[01:37:15] <DontKnwMuch> try some other board if you have it around[01:37:17] <RoyK> I have three servers showing issues with that board[01:37:19] *** dijenerate has quit IRC[01:37:31] <blues> I've got an asus board i can try[01:37:41] <RoyK> richlowe: and 9201 does not show those issues, at least not that badly[01:37:44] <blues> x58 chipset though.[01:37:54] *** jwit has quit IRC[01:38:03] <RoyK> richlowe: please explain what is so wierd with my config[01:38:34] *** gea__ has quit IRC[01:38:45] <DontKnwMuch> try this one. but I had good experience with the simplest of the boards to do the flashing with, same pal error, when I tried on some intel server board, but it went ok on another[01:39:28] <DontKnwMuch> a p33 or somehting I think was ok[01:40:14] <RoyK> richlowe: if you can't explain what's so "weird" about my config, then please stfu[01:40:52] <DontKnwMuch> royk: what issues, just interested, I have some 1068 I wanted to use..[01:42:48] *** DrLou has quit IRC[01:43:16] *** jwit has joined #openindiana[01:43:46] <RoyK> DontKnwMuch: massive i/o errors after some time, especially during high load[01:44:04] <RoyK> richlowe: did you fall asleep or didn't you have any answer?[01:44:34] *** kimc has quit IRC[01:46:48] <RoyK> richlowe: I use desktop driver, ok, but those drives are used by thumpers and all sorts of "enterprise" stuff - the problems we see are related to the 1068 controllers and/or the drivers for those[01:46:57] <blues> DontKnwMuch: tried another board, same problem[01:47:03] <blues> Failed to initialize PAL[01:47:14] <RoyK> richlowe: enterprise drives are just a firmware upgrade to desktop drives and you know it[01:48:09] * RoyK is tired of noobs that beleive anything labelled "enterprise" is built otherwise than things built for the desktop[01:48:16] <tomww> how does firmware make hardwar last longer?[01:49:13] <RoyK> tomww: it doesn't, but for some makes, like WD, it adds TLER, which makes the drives tell a sector is bad earlier instead of retrying too much[01:51:28] <RoyK> tomww: the deafault for desktop WD drives is to retry a lot before returning an error, which is rather bad if you have it in a RAID[01:51:45] <tsoome> "The ability to change the TLER option on Western Digital Desktop and Enterprise hard drives..."[01:52:16] <RoyK> tsoome: ...disappeared a year ago or so[01:52:32] <tsoome> that part was not important;)[01:52:32] <DontKnwMuch> blues: strange thing...[01:53:04] <RoyK> but FUD from richlowe about using desktop drives fucking up the system is nonsense[01:53:15] <tsoome> TLER is not something what does identify enterprise;)[01:53:20] *** jwit_ has joined #openindiana[01:53:27] <blues> this board has EFI...should i try the EFI installer maybe?[01:53:34] *** jwit has quit IRC[01:53:49] <RoyK> tsoome: 'cept it only exists on WD's so-called enterprise drives[01:54:47] <tsoome> uhm.... well, i dont know, that sentence does have *and* between words desktop and enterprise;)[01:54:48] <DontKnwMuch> blues: yes, I read somewhere that EFI might work in this case[01:55:07] <DontKnwMuch> blues: EFI flasher that is[01:55:09] <tsoome> but i really am no expert about that:D[01:56:49] <RoyK> richlowe: could you explain what's so weird about my setup, please?[01:56:54] *** sergefonville has left #openindiana[01:57:29] <tsoome> http://forum.synology.com/wiki/index.php/Differences_between_an_Enterprise-Class_HDD_and_a_Desktop-Class_HDD[01:58:17] <RoyK> tsoome: and then, the thumper ships with deskstars........[01:58:28] <tsoome> well, tbh, im sure in some cases the line between those 2 classes may be really thin, but I wouldnt tell that all those drives are the same;)[01:58:37] *** axisys has quit IRC[01:58:46] <tsoome> well, drive inside thumper does not make it enterprise[01:58:49] <RoyK> tsoome: the whole idiocity about "thou shalt use enterprise drives" goes a long way[01:59:05] <tsoome> thumper itself is noway near enterprise[01:59:14] <RoyK> fuck this[01:59:18] <tsoome> :)[01:59:21] <RoyK> desktop drives work[01:59:26] <RoyK> for all systems[01:59:27] <tsoome> sure[01:59:41] <RoyK> but the issues I have with 1068 are not related to the drives[01:59:54] <tsoome> till you have issue when specific firmware does not work with your HBA[02:00:02] <RoyK> even though some besserwissers may want to say so[02:00:29] <RoyK> not an issue - have tried several - on several servers[02:00:47] <RoyK> fuck it - I hate it when people try to blame problems on hardware[02:01:04] <tsoome> firmware is just software[02:01:17] <tsoome> and there have been such issues before[02:01:39] <RoyK> and, sorry to point out that you haven't gotten the point, that's not the issue either[02:01:43] <tsoome> I remember intalling disk firmwares on A5000 array[02:01:57] * RoyK steals tsoome's beer[02:02:01] <tsoome> the issue was that under high load the disk did block the IO[02:02:16] <tsoome> i was pretty amused at that time:)[02:02:51] <RoyK> tsoome: do you really think I haven't tried to flash up the controllers to fix this?[02:03:09] <RoyK> it affects all drives I've tested[02:03:20] <RoyK> on LSI 1068[02:03:27] <tsoome> another one - they say that in many cases, if you see checksum errors on hdd or ssd, its really caused by faulty disk firmware:)[02:03:46] *** xmikus01 has quit IRC[02:03:46] <RoyK> fuck offf[02:04:43] <DontKnwMuch> royk: what disk do you use, so I know what to avoid ;)[02:04:58] <tsoome> well, the reality is, in such case you just dont know. it can be bug in driver, it can be bug in hba firmware, it can be hdd. HDD is easiest to exclude by switching brands.[02:05:08] <tsoome> rest is kinda crap:([02:05:34] <tsoome> thats the case when its handy to have actual vendor support:([02:06:54] <DanaG> Apple advertises their Time Capsule as having "server-grade" hard drives.[02:06:58] <tsoome> hehe, i found funny metric from the intel doc referenced in that wiki entry about enterprise versus desktop drives:D[02:07:12] <DanaG> http://www.zdnet.com/blog/hardware/hey-apple-wheres-that-server-grade-time-capsule-hard-drive-you-promised-us/13451[02:07:15] <tsoome> enterprise disks have leds, desktop ones dont:D[02:07:19] <tsoome> lol[02:07:47] <joffe> very important[02:07:57] <joffe> LEDs increase MTBF by 2 million hours[02:08:06] <joffe> if they blink, 3mill[02:08:12] *** realjoe has joined #openindiana[02:08:17] <tsoome> well, its from 2008, so no need to pay too much attention, but still funny:)[02:08:34] <realjoe> hi, do some dvb tuners work with openindiana?[02:08:46] <tsoome> but there are still some useful bits.[02:09:57] *** jwit_ is now known as jwit[02:10:20] <tsoome> hm, what they mean with spindle anchored in both ends versus one end?[02:10:53] <tsoome> ah, ok, got it...[02:12:45] <DontKnwMuch> I opened one dead samsung 500gb I beleive drive, it had spindle anchoer in both ends... it its price when new was 36eur or so...[02:13:38] <tsoome> ya well, that intel doc is from 2008, many technical details were probably changed on vendors even at that time, not to speak about today[02:14:36] <realjoe> I am running open indiana within virtualbox, but the package manager and update manager and gparted wont open, what can I do?[02:15:01] *** Botanic has joined #openindiana[02:15:06] <tsoome> give it more ram?[02:15:27] <tsoome> how much it got?[02:16:05] <DanaG> realjoe: try this in console: sudo passwd root[02:16:17] <DanaG> I'm guessing it may be the root-password-expired bug.[02:17:13] <realjoe> worked! thanks![02:17:33] <blues> i think i'm ready to cry now[02:17:46] <tsoome> blues: go ahead:)[02:18:00] <tsoome> why?[02:18:07] <realjoe> Is it worth to install the VBoxSolarisAdditions.pkg ?[02:18:28] <blues> tries controller in another box... that has been working fine... now VGA-LED is on the board and i'm not getting video[02:20:19] <DanaG> say, does OI support pools with GPT partition tables instead of MBR?[02:20:35] <tsoome> RoyK: was that the issue about disks giving errors with that specific HBA?[02:20:53] <tsoome> DanaG: only on data pools, not rpool[02:21:23] <tsoome> in fact, thats default if you create pool from entire disk setup[02:23:59] <DontKnwMuch> blues: vga-led?[02:24:10] <blues> little LED on the mobo labeled VGA[02:24:20] <blues> looking for mobo manual now to intepret said LED[02:24:25] <DontKnwMuch> you have onboard vga?[02:24:43] <DontKnwMuch> and plugged th 1068 in 16x slot?[02:25:01] <blues> i've removed all cards[02:25:13] <blues> and plugged a vid card into x16 slot[02:25:49] <DontKnwMuch> oh ... try to reset bios with the jumper...[02:25:56] <blues> i have[02:26:01] <DanaG> hmm, try looking for "init display first"?[02:26:19] <DanaG> It may be trying to use the non-VGA card as boot graphics, which of course won't work.[02:27:16] <blues> right..except i've got nothing... i can't see post.. and now have no cards installed at all. It was working fine btw... first time i booted with controller in x16 slot + onboard video... then i changed some bios settings (boot order) and reset...and nothing[02:28:30] <DanaG> I mean, you'd boot without card, to change option, then re-add the card.[02:28:42] <blues> lmao...and asus's download site doesn't offer an English manual[02:31:13] <mnaser> blues: don't be a wuss[02:31:13] <mnaser> yo[02:31:14] <mnaser> u[02:31:30] <mnaser> you don't need no texts for manual! just look at the screenshots.[02:32:22] <DontKnwMuch> a manual? what is that?[02:33:41] <DanaG> hmm, is there any way to install into an existing rpool?[02:33:43] <DontKnwMuch> I would just pull everything out of the board, ram, cables, front panel connector, all, and just put one ram, power and a screwdriver to the power on pin. It damn shoudl work if it did a few minutes ago...[02:36:08] <mnaser> pro tip don't move around laptop when its burning a cd[02:39:48] <DanaG> https://defect.opensolaris.org/bz/show_bug.cgi?id=5840[02:44:21] *** axisys has joined #openindiana[02:44:26] <DanaG> hmm, maybe I should just bite the bullet and install the OS to the mirror.[02:44:48] <DanaG> Is there any way to specify pool version, while doing so?[02:47:47] <DanaG> Eh, to heck with it, I'll just go all-out OpenSolaris, and if I change my mind, I'll re-copy from the backup drive.[02:48:28] <DanaG> And since USB is so slow, I might as well just have the OS and the data in the same mirror pool.[02:50:46] *** freedomrun has quit IRC[02:52:25] <realjoe> I would like to add this repo: http://staticdev.uk.openindiana.org:10002/en/index.shtml how can I do that?[02:55:33] <blues> this is insane[02:56:47] *** axisys has quit IRC[02:58:32] *** axisys has joined #openindiana[02:59:05] *** DanaG has quit IRC[03:00:29] <herzen> realjoe: pkg set-publisher -g http://staticdev.uk.openindiana.org:10002 oi-sfe[03:00:44] <realjoe> thanks[03:00:57] <herzen> but there's also another publisher that uses port 100003[03:01:07] <herzen> *10003*[03:01:55] <herzen> the main differences between the two is that the latter uses modern package names and doesn't have multimedia packages, other than mpd without any encumbered coders, like mp3[03:02:00] <realjoe> ok, before I do something wrong: I thought this is something like the upcomming official packages[03:02:18] <herzen> 10003 has the upcoming official packages[03:02:31] <herzen> 10002 was more experimental[03:02:35] <realjoe> ah ok[03:02:47] *** axisys has quit IRC[03:03:11] <realjoe> then I will use 10003, I am used to apt and launchpad, how do I know what repos are available and which I can trust?[03:03:20] <herzen> we're going to set up a separate repo for packages using multimedia codecs that have patent issues[03:03:32] *** axisys has joined #openindiana[03:03:50] <herzen> you can trust me on those two repos :-)[03:04:21] <herzen> otherwise, I guess be suspicious of anything that's not mentioned in the OI wiki[03:05:42] <realjoe> it says it contains packages for ebuild, not oi-sfe[03:06:07] <herzen> what's ebuild?[03:06:36] <realjoe> oh, i missed some letter: it says: sfebuild[03:07:20] <herzen> why we haven't been publicizing it is that we still have to sort out the issue of what to use for copyright information[03:08:05] <herzen> yeah, sfebuild is the name of the repo with this URL: http://staticdev.uk.openindiana.org:10003[03:09:02] <herzen> also, it appears that the only page on the Wiki saying what repos there are is written in German[03:09:35] <realjoe> no problem, as I am german ^^[03:10:00] <herzen> you're in luck, then ^^[03:10:13] <realjoe> still your command wont add the repo, because of the difference in sfebuild and oi-sfe[03:10:23] <herzen> sorry[03:10:46] <herzen> pkg set-publisher -g http://staticdev.uk.openindiana.org:10003 sfebuild[03:10:47] <realjoe> is this normal?[03:10:49] <herzen> that should work[03:10:53] <realjoe> ah ok[03:11:01] <herzen> with pfexec, of course[03:11:14] *** master_of_master has quit IRC[03:11:44] <herzen> what packages are you interested in, if I may ask?[03:11:49] <realjoe> hey thanks, it was added[03:13:02] *** master_of_master has joined #openindiana[03:13:11] <herzen> please let us know, preferably on #pkgbuild, if you run into any problems[03:13:39] <herzen> the plan is to have the SFE contributors support those packages, not the OI devs[03:13:52] <realjoe> do you mean me?[03:14:01] <herzen> yes[03:14:33] <realjoe> sorry I dont know the difference between SFE and OI[03:14:37] <herzen> #pkgbuild is the SFE channel[03:14:47] <herzen> yeah, that should be explained in the Wiki[03:15:41] <herzen> SFE is a project that started with OpenSolaris, to provide specs for packages not provided by OpenSolaris itself[03:15:58] <herzen> but it itself does not provide binary packages or an IPS repo[03:16:19] <herzen> Oracle killed OpenSolaris, which is why OI was started[03:17:13] <herzen> so OI is hosting packages built from SFE specs on an IPS repo, but it does not have the manpower to support them itself[03:17:23] <realjoe> I see[03:17:56] <herzen> that's the easiest way for OI to add binary packages of stuff it doesn't have, like mplayer or mpd[03:18:21] <realjoe> can linux drivers for dvb cards be used with OI?[03:18:47] <herzen> I would doubt it, but that kind of thing is not really my line[03:19:06] <herzen> OI uses the Solaris/Illumos kernel[03:19:13] <realjoe> I see[03:19:26] <herzen> and I think drivers are usually kernel-specific[03:34:42] <blues> could i have fried this damn board putting the intel SAS controller in the x16 pcie port?[03:36:06] *** DanaG has joined #openindiana[03:36:30] <DanaG> Okay, so I have it installed on one internal drive, and I started randomly wondering: does napp-it use samba daemon, or in-kernel cifs?[03:38:15] *** ivo_ has quit IRC[03:40:26] *** ThomasB2k has quit IRC[03:40:35] *** dijenerate has joined #openindiana[03:48:37] *** movement has quit IRC[03:51:01] *** realjoe has left #openindiana[03:56:30] <DanaG> I'm trying to mirror my rpool, and I keep getting:[03:57:01] <DanaG> fmthard: Partition 2 specifies the full disk and is not equal full size of the disk. The full disk capacity is .... sectors.[03:57:23] <DanaG> Is that just a warning?[03:57:57] <DanaG> I see... the second drive is precisely 1 cylinder smaller than the first![03:58:11] <DanaG> Should I just redo the install onto the second drive?[03:58:29] <herzen> that's the only solution I can think of[03:58:40] <herzen> are the drives different makes/models?[03:58:51] <DanaG> Nope, both are Samsung F4, ordered about 1 week apart.[03:58:54] <DanaG> 2TB.[03:59:17] <herzen> well, this makes me not want to buy Samsung drives. not that I ever have[03:59:56] <DanaG> And I don't know offhand which one is bigger.[04:00:04] <herzen> zfs is fine with mirroring devices of different sizes. it's just that the capacity used has to be the minimum of that of all the drives, obviously[04:00:12] <DanaG> A workaround would be for the installer to leave a bit of space free at the end.[04:00:23] <DanaG> Oh, so that'll do.[04:00:36] <DanaG> The error was throwing me.. or rather, the lack of notation of whether that's an error or just a warning.[04:00:42] *** movement has joined #openindiana[04:01:39] <DanaG> Will I have to redo the install, or is it okay to just add the "smaller" drive to the mirror?[04:01:50] <DanaG> I'll have to check the serial numbers.[04:02:02] <herzen> I don't know if that's OK[04:02:19] <herzen> I guess you can try seeing if zfs lets you get away with that[04:02:35] <herzen> it's a fresh install apparently after all, so you have little to loose if you botch it[04:02:41] <DanaG> eh, it's easier to just redo the install.[04:02:56] <DanaG> er, rather, I'm just indecisive.[04:03:13] <DanaG> I think I got an error when I tried to add to the pool... don't remmber what the error was.[04:03:20] <herzen> If I knew more about hard drives, I could probably tell you how to verify that the two drives actually have a different number of sectors[04:03:28] <DanaG> So I'll just redo with smaller drive.[04:04:39] <herzen> it could be you used a sector on one drive for something and you didn't on the other drive[04:04:52] <DanaG> I'll try hdparm, if it exists.[04:04:53] <herzen> I'm not an expert here, so sorry I can't help you more[04:06:19] <DanaG> hmm, so now that I'm using the hard drives as boot volumes, I'm not sure what I want to use the internal USB port for.[04:08:00] <DanaG> I wish HP would let you specify the IPMI card as primary graphics, while leaving the HD4200 enabled.[04:10:33] <DanaG> yeah, I'm just redoing the install.[04:11:02] <DanaG> Oopsie, now I've lost track of which drive was the smaller one.[04:11:13] <DanaG> It seems t1 is somehow the smaller one again. I thought I swapped them.[04:13:38] <DanaG> Swapped them back. I have suspicions that it's not really the drives doing it.[04:14:13] <DanaG> So I'll install with, say, 100MB left unpartitioned at the end of the disk.[04:14:20] <DanaG> That should cover 1 cylinder, right?[04:17:45] <herzen> I guess. like I said, I'm not an expert[04:17:58] <DanaG> nope, I swapped them back, and one really is smaller. I'll still leave a bit of space, though.[04:21:30] *** nikolam has quit IRC[04:24:25] *** LeftWing__ has joined #openindiana[04:25:25] *** LeftWing has quit IRC[04:26:08] <DanaG> bah, it's not putting the partition beginning at multiples of 4K.[04:28:00] <DanaG> I don't want to change ashift; I just want to make sure the partition starts at a multiple of 4K.[04:32:01] *** InTheWings has quit IRC[04:44:11] <DanaG> Cool, I figured out how to align the start of the partition.[04:44:18] <DanaG> Without the zpool hack.[04:44:34] <DanaG> Here's how: I created a dummy 0.1GB Extended partition in the installer, in front of the ZFS.[04:45:38] <DanaG> So now the sun partition begins at 209668, which is divisible by 4 (four 512s in 4K).[04:47:10] *** LeftWing__ is now known as LeftWing[04:50:27] *** floogy has joined #openindiana[04:52:18] <mnaser> DanaG: aligning an ssd fs?[04:52:51] <DanaG> No, a 4K hard drive.[05:00:52] <DanaG> Yeah, so it looks like ashift change would break things, so all I did was align the start.[05:01:46] <DanaG> heh, one downside/upside (not sure which) of not using the onboard NIC, is that I don't get the LAN LED (right next to HDD LED) lighting up.[05:03:18] *** floogy has quit IRC[05:04:03] *** Loafamotive has joined #openindiana[05:28:40] <DanaG> okay, I tried to use dd to copy the partition table from one drive to the other... and it didn't work.[05:29:25] <DanaG> But I was using s0, s1, and such.[05:29:37] <DanaG> What's the equivalent of /dev/sda and /dev/sdb (bare drives)?[05:43:36] <DanaG> okay, so I created a new partition table, and used c1t1d0s2 to get "whole disk" from source drive.[05:43:48] <DanaG> Now, when I try to dd that to the target drive, I get "read-only filesystem"[05:45:25] <DanaG> ah, I see.. now it's bare at c1t1d0[05:46:20] <DanaG> nope, that didn't do it.[05:48:10] <DanaG> say, what was that oi-cbe (or whateever it was) repo?[05:48:22] <DanaG> er, silly me, google.[05:48:44] <DanaG> oi-sfe[05:49:10] <DanaG> Nope, that's not it, either.[05:49:52] <DanaG> http://forums.nekochan.net/viewtopic.php?f=17&t=16725293&view=unread[05:58:58] <DanaG> Okay, why is it so hard to DD to the beginning of an unpartitioned disk?[06:16:12] *** tsukasa has quit IRC[06:29:29] *** tsukasa has joined #openindiana[06:37:11] *** redgone has quit IRC[07:02:59] *** Cobi has quit IRC[07:11:52] *** Cobi has joined #openindiana[07:14:38] *** freedomrun has joined #openindiana[07:48:07] <DanaG> say, does napp-it use samba daemon, or kernel CIFS?[08:08:38] <DanaG> looks like it uses kernel cifs... but failed when trying to svcadm enable.[08:15:21] <DanaG> Anyway, OI installer created homes case-sensitive... I'd like to make them insensitive. Is there a programmatic way to recreate it, with case-sensitivity=mixed?[08:15:37] <DanaG> Note that I don't care about wiping out data -- there should be no significant data there.[08:24:52] *** Naresh``` has joined #openindiana[08:25:55] <DanaG> ah, nevermind... I can just do some "zfs get" and "cut" trickery, to see only what's been specifically set.[08:27:45] <DanaG> And how can a property have "none" as a source?[08:27:57] <DanaG> the options are: local,default,inherited,temporary,received,none[08:31:14] *** Naresh``` is now known as Naresh[08:31:17] *** Naresh has joined #openindiana[08:45:22] <DanaG> hmm, I removed /export/home and thus /export/home/dana, and even after recreating it, my home dir doesn't automount.[08:47:16] <DanaG> hmm, I'll try rebooting.[09:04:45] <DanaG> great, some combination of that, and using napp-it, made my root password not work.[09:05:04] <DanaG> Nor does my user password work, since I deleted and recreated myself.[09:11:42] *** Naresh has quit IRC[09:12:58] *** Naresh has joined #openindiana[09:16:41] *** sivanov_ has joined #openindiana[09:17:26] <dkeav> grah![09:24:59] <DanaG> Got my passwords reset, but now rpool/ROOT/openindiana has mountpoint=/[09:25:03] <DanaG> er, it's /b[09:25:13] <DanaG> What's it on your system? Just do zfs list.[09:25:56] <alanc> rpool/ROOT/opensolaris 70.2M 174G 10.7G /[09:26:10] <dkeav> rpool1/ROOT/openindiana 3.16G 27.1G 3.07G /[09:26:12] <DanaG> hmm, are there any other things under ROOT with the same mountpoint?[09:26:25] <DanaG> I tried to reset it to '/', and it wants to unmount the existing root.[09:26:29] <alanc> all my BE's have mountpoint /[09:26:48] <DanaG> I was following this: http://comments.gmane.org/gmane.os.openindiana.general/1236[09:27:09] <alanc> rpool/ROOT/opensolaris-117, rpool/ROOT/opensolaris-134, rpool/ROOT/snv_151a, rpool/ROOT/snv_168-X169 - all /[09:28:10] <alanc> I really should clean up some of these BE's for 2 year old builds at some point[09:28:38] <dkeav> yea[09:28:43] <dkeav> like now[09:28:58] <DanaG> So now I can't seem to reset that mountpoint.[09:29:30] <DanaG> trying to log in as root: roles can only be assumed by authorized users[09:29:47] <DanaG> trying to log in as me: it just exits immediately.[09:34:04] <DanaG> I see, root is not a real user.[09:34:12] <DanaG> So I just need to figure out why my own user isn't working.[09:40:27] <DanaG> okay, looks like I need to re-enable automounting /export[09:40:54] <dkeav> re-enable?[09:40:58] <dkeav> wtf man[09:41:54] <dkeav> k i know i'm fairly inhebriated but tell me you are not learning unix in a live environment?[09:44:28] <DanaG> My own server, with no useful data on it yet.[09:44:42] <dkeav> whewww[09:44:57] <DanaG> I'm plenty familiar with Linux; not familiar at all with Solaris.[09:45:08] <DanaG> All I was trying to do was to enable casesensitivity=mixed on /export[09:46:05] <DanaG> To me, "production" is a better word. Live is ambigous... livecd?[09:47:11] <dkeav> k lets sayd production[09:47:16] <dkeav> for the sake of arguement[09:47:17] <dkeav> :D[09:48:08] <DanaG> I did get what you meant, after a little while, though.[09:49:13] <dkeav> i was just hoping you were learning unix in a suitable environment[09:50:13] <dkeav> i kinda grew up in unix so meh, but helping people get acclimated i tend to let them use VM's to pound on for awhile[09:50:36] <DanaG> Okay, so now all I need is to recreate my user and have it create the correct automount setup.[09:51:05] <DanaG> All I see on useradd would let me have home dir be in /export/home/dana, OR /home/dana... but it won't create /export/home/dana if I specify /home/dana, will it?[09:51:35] <dkeav> its a symlink[09:52:20] <DanaG> Anyway, it created /home/dana, but no /export/home/dana[09:52:23] <DanaG> And no zfs data set[09:54:36] *** Naresh has quit IRC[09:56:04] *** Naresh has joined #openindiana[09:58:52] *** adama has quit IRC[10:00:59] <DanaG> Isn't useradd supposed to create a dataset?[10:01:04] <DanaG> It sure isn't doing that for me![10:02:31] <dkeav> iirc it only creates a new instance in /etc/passwd|/etc/shaddow and /etc/user_attr[10:02:45] <dkeav> but i'm fairly drunk and could be remembering wrong[10:05:16] <dkeav> what you are wanting was a CR for zfs at one point[10:05:24] <dkeav> but i dunno if it ever found its way in[10:08:22] <dkeav> errr actually looks like that CR is still in pipe[10:08:35] <DanaG> CR? I don't know what that stands for.[10:08:59] <dkeav> code review[10:10:06] *** syoyo_ has joined #openindiana[10:10:30] <DanaG> I see... I had to create /export/home/dana manually, and then set password, and then log in.[10:10:47] <DanaG> er, wait... no, I'm in /[10:12:14] <dkeav> heh[10:12:22] <dkeav> on that note, i'm gonna go sleep one off[10:12:24] <dkeav> laters[10:19:18] <DanaG> ah, reboot out of single fixed it.[10:23:27] <DanaG> Still screwed up, though... .profile isn't getting read![10:23:59] <DanaG> ugh, bed is a good idea.[10:25:22] *** xmikus01 has joined #openindiana[10:26:14] <dkeav> aye[10:26:17] <dkeav> http://www.youtube.com/watch?v=Xr9_e4ySRYA[10:26:22] <dkeav> to help you sleep[10:26:28] <dkeav> nighty night[10:26:29] <DanaG> I think I'll just say "screw it", and reinstall.[10:26:54] <DanaG> I just wish the danged installer had an option for case sensitivity![10:27:02] <DanaG> Is the installer a script, or a binary?[10:29:44] <herzen> case-insensitivity is utterly non-Unix: Unix isn't Lisp, after all[10:30:09] <herzen> so I don't think it's occurred to anyone to consider such an option[10:31:29] <herzen> I'm surprised that zfs even provides for that[10:32:00] <DanaG> casesensitivity=mixed... lets it respond to case-insensitive requests.[10:32:09] <DanaG> Useful if you want to share the whole dang home dir via samba.[10:34:14] <DanaG> Unfortunately, you can't change that casesensitivity value after creation.[10:34:28] <herzen> oh. I haven't had to use Samba for several years[10:34:51] <herzen> I'm surprised you can't change that after creation. that doesn't seem right to me[10:35:01] <herzen> anyway, on a serious note, is this for a home network?[10:35:14] *** gea__ has joined #openindiana[10:36:31] <herzen> because when I have to make Windows an Solaris/OI share files, I now use nfs[10:37:04] <herzen> Windows 7 comes with a nfs3 client, and you can download one from Microsoft for Windows XP[10:44:56] <DanaG> How does it deal with case-sensitivity on nfs?[10:45:10] <DanaG> And does OS X do nfs?[10:46:02] <herzen> I would imagine there's a good chance that OS X does nfs, since it's based on freeBSD, I believe[10:46:16] <DanaG> http://www.opensolaris.org/jive/thread.jspa?messageID=316464&tstart=0[10:46:53] <herzen> and I have never had problems with case sensitivity with a Windows 7 client interacting with a Solaris nfs server[10:47:25] <herzen> but I hardly use Windows at all, so maybe something I don't know about doesn't work in that regard[10:48:37] <DanaG> Hmm, another possible way I can set it case-sensitive: do it while still booted from the livecd, and make damn sure to get the permissions right.[10:48:52] <DanaG> Can you take snapshots of changes to the existence of pools?[10:48:58] <herzen> it's just not straightforward how to set up nfs to serve Windows, since the Solaris default is nfs version 4, but Microsoft has only implemented version 3[10:49:43] <tsoome> why that should matter?[10:49:57] <tsoome> solaris nfs server supports both v3 and v4[10:50:23] <herzen> yes, but zfs assumes you want to use v4, as far as I understand[10:50:54] <herzen> so when I've shared with Windows, I have bypassed zfs to create the nfs shares[10:51:04] <herzen> maybe there's a better way to do that I don't know about[10:51:06] <tsoome> if you share nfs, its shared with both v3 and v4[10:51:51] <herzen> OK, I guess I misunderstood something somewhere[10:52:20] <DanaG> Does nfs do security?[10:52:22] <herzen> I haven't played around with sharing nfs with Windows that much; I just prefer it to Samba[10:52:56] <herzen> anyway, it's time for me to go to bed now, too. see you all later[10:53:09] <tsoome> DanaG: yes it does. if you configure it. if you dont, it does not.[10:56:55] <DanaG> I've actually found the case-sensitivity in Linux sometimes annoying.[10:57:13] <DanaG> Like, chromium downloads to "downloads" dir, whereas everything else downloads to "Downloads" dir.[10:58:14] <tsoome> cant it be configured?[10:58:19] <DanaG> Yeah.[10:58:30] <DanaG> But I've also had similar issues with torrents, I seem to recall.[11:04:24] <DanaG> At this point, I've been up so long that I'll be less messed-up if I don't sleep.[11:04:27] <DanaG> well, maybe.[11:37:02] *** freedomrun_ has joined #openindiana[11:37:33] *** freedomrun has quit IRC[11:51:34] *** DanaG has quit IRC[12:00:32] *** land0 has joined #openindiana[12:03:08] *** bens1 has joined #openindiana[12:03:14] *** stoxx has quit IRC[12:16:25] *** stoxx has joined #openindiana[12:19:17] *** land0 has quit IRC[12:38:01] <sivanov_> unfortunately, windows7 nfs client didnt support utf-8 last time ive checked[13:03:33] *** ivo_ has joined #openindiana[13:45:00] *** sivanov_ has quit IRC[13:47:33] *** descipher_ has joined #openindiana[13:49:52] *** descipher has quit IRC[13:49:59] *** descipher_ is now known as descipher[14:13:25] <gea__> Vmware-tools (included in ESXi 4.1 ) are installed per default with a perl-script[14:13:27] <gea__> This installer is not running with new current perl in OpenIndiana 151 (only a lot of perl errors, no problem with OI 148 or SE 11 build 151)[14:13:29] <gea__> Most virtualization is done with vmware so i think this is absolute essential to fix it.[14:13:31] <gea__> Any ideas how to fix or am i the only who care about this problem (I filed a bug report to OpenIndiana but it was rejected as not being a OI problem)[14:13:33] <gea__> I have had similar problems with my perl installer script called from CLI. i could fix it there by removing use strict. Similar scripts are running from CGI without problem[14:13:34] <gea__> so it seems a CLI problem only.[14:18:00] *** sivanov_ has joined #openindiana[14:25:26] <sivanov_> does pciconf irq settings stay across reboots?[14:25:29] <sivanov_> *do[14:27:25] *** InTheWings has joined #openindiana[14:59:35] *** adama has joined #openindiana[15:44:43] *** syoyo_ has quit IRC[15:46:38] <sivanov_> initiator:FreebSD 8.2 qle 2460 http://www.pastie.org/2191602[15:47:20] <sivanov_> local: http://www.pastie.org/2192059[15:48:30] <sivanov_> initiator gets 200MB/s less read speed, interesting, why[16:14:37] <sivanov_> according this diagram of the motherboard all pci devices are wired to SB which is wired to cpu0, is it logical to put their interrupt handlers on cpu#0??[16:37:13] *** hsp has joined #openindiana[16:40:55] *** redgone has joined #openindiana[16:42:09] *** sivanov_ has quit IRC[17:07:10] *** nikolam has joined #openindiana[17:24:40] *** miine has quit IRC[17:31:20] *** miine has joined #openindiana[17:44:22] *** syoyo has joined #openindiana[17:56:23] *** wonslung has quit IRC[18:19:25] *** xmikus01 has quit IRC[18:23:34] <DontKnwMuch> it is using kernel cifs[18:34:27] *** freedomrun has joined #openindiana[18:35:44] *** freedomrun_ has quit IRC[18:45:59] *** lblume has quit IRC[18:47:22] *** lblume has joined #openindiana[18:58:06] *** lblume has quit IRC[18:59:31] *** DanaG has joined #openindiana[19:00:14] *** lblume has joined #openindiana[19:15:06] <DanaG> Weird.. my zfs smb share is rejecting my attempts to log in as 'dana'.[19:15:21] <DanaG> It's saying user 'guest' is not allowed -- so it's totally ignoring the username![19:15:47] <tsoome> you did change the password after adding pam module?[19:16:29] <tsoome> or are you using AD?[19:16:38] *** ivo_ is now known as anAnunaki[19:16:49] <DanaG> No AD. Oh, wait, I see... I need to re-add that pam module, because I reinstalled.[19:17:23] <DanaG> Thanks.[19:19:57] <DanaG> Now I get: "dana access denied: RPC only", and "dana access denied: guest disabled".[19:23:09] <DanaG> and above that is "ad lookup of dana@ failed"[19:28:07] <DanaG> I also have idmap winuser:* unixuser:*[19:33:24] *** yalu_ has joined #openindiana[19:33:25] <DanaG> "share: failed to remove ACL object: c$: no such file or directory[19:33:27] <DanaG> "[19:36:23] *** yalu has quit IRC[19:36:51] <DanaG> Caution - Do not use name-based mapping rules when in workgroup mode.[19:36:53] <DanaG> ah.[19:36:55] <DanaG> http://forums.oracle.com/forums/thread.jspa?threadID=2243032&tstart=0[19:38:13] *** systemonkey has joined #openindiana[19:38:43] *** syoyo has quit IRC[19:39:10] <systemonkey> hi guys, my rpool dissappeared... and I'm in panic mode...[19:39:27] <tsoome> why it did disappear?[19:39:52] <DanaG> Say, why doesn't "format" show a hotplugged disk?[19:40:06] <systemonkey> hi tsoome, don't know. when I run zpool status, it says no pools available.[19:40:24] <systemonkey> all vdevs seems to be online tho[19:40:26] <tsoome> what did you do before it disappeared?[19:40:39] <systemonkey> nothing. I just a call from office saying that the system is down[19:41:04] *** lblume1 has joined #openindiana[19:41:11] <tsoome> DanaG: ru cfgadm -al and then cfgadm -c configure on disk which got unconfigured state[19:41:18] <systemonkey> so when I saw the error of the application, it says the data is missing. so naturally I checked the pool and it is no longer there[19:41:33] <tsoome> mirrored rpool?[19:41:37] *** lblume has quit IRC[19:41:48] *** bens1 has quit IRC[19:41:58] <systemonkey> no it wasn't mirred rpool[19:42:03] <tsoome> ou.[19:42:13] <systemonkey> am I screwed?[19:42:16] <tsoome> well, dmesg does not show anything?[19:42:27] <DanaG> hmm, it doesn't appear there.[19:42:43] <tsoome> maybe, maybe not. if the disk is really dead....[19:42:56] <DanaG> If rpool is mirrored, you shouldn't be able to boot at all, right?[19:43:05] <DanaG> er, if it's missing/[19:43:17] <DanaG> if you're on the live cd/dvd, try zpool import[19:44:22] <DanaG> I see... for my issue, cfgadm on the sata port, not the disk.[19:44:22] <systemonkey> unfortunately I'm not near the system...[19:44:23] <tsoome> in some cases the HBA has "disconneced" the drive. basically you can try reboot, if that does not help, boot dvd and check[19:44:38] <tsoome> DanaG: aye:P[19:45:02] <DanaG> What would it take to make it automatically do that?[19:45:35] <systemonkey> would recreating the pool using the same vdevs bring the databack?[19:46:00] <tsoome> there is no point to automate it really, as you will need to identify disk and create/add pool anyhow[19:46:04] <DanaG> I'd try import, first. what does "zpool import" show?[19:46:16] <systemonkey> no available pools...[19:46:18] <DanaG> tsoome: right now, the disk I hotplugged is one with NTFS. is there ant read support, or none?[19:46:19] <tsoome> no, dont destroy/create anyhing[19:46:21] <DanaG> anu[19:46:26] <DanaG> any (dang, can't type)[19:46:59] <tsoome> ntfsng or something, i have no idea about exact packgage name or details[19:47:38] <DanaG> Would it be faster to do that, or to copy over gigabit (from, Intel to Intel nics) with client in Windows?[19:48:00] <tsoome> depends how much you need to copy i guess[19:48:00] <DanaG> ah, ntfsprogs.[19:48:53] <DanaG> Doesn't contain a "mount"[19:49:05] <tsoome> mount -F ntfs?[19:49:47] <DanaG> mount: operation not applicable to fstype nfs[19:50:08] <tsoome> nTfs:P[19:50:13] *** nikolam has quit IRC[19:50:23] <DanaG> sorry, typo was not on running, but on typing here[19:50:27] <systemonkey> well... this is weird. now I'm not getting any output after I run zpool import[19:50:34] <DanaG> (I'm using microserver vKVM -- no copy and paste)[19:50:38] <tsoome> zpool list?[19:50:59] <systemonkey> no pools available when running zpool list[19:51:18] <tsoome> you did reboot?[19:51:22] <systemonkey> yes[19:51:31] <DanaG> How did it boot without rpool?[19:51:36] <tsoome> and it did boot but still does not list rpool?[19:51:54] <tsoome> iostat -En ?[19:51:55] <DanaG> Are you running as admin (su -, or sudo -i)?[19:52:45] <systemonkey> no the other admin decided to use zfs-fuse on this particular machine. So I have one machine running openindiana and this one is running on zfs-fuse on centos[19:53:21] <tsoome> cant help about that fuse. never used it, 0 experience[19:53:32] <systemonkey> ouie...[19:54:08] *** raichoo has joined #openindiana[19:54:19] <DanaG> Rather, I meant as root. It won't show anything if the command doesn't have privileges.[19:54:24] <raichoo> hi folks![19:54:29] <raichoo> hi folks[19:55:19] *** DanaG has quit IRC[20:00:47] <raichoo> Whoops, double post ^^[20:11:28] *** ThomasB2k has joined #openindiana[20:13:04] <dkeav> don't let it happen again[20:15:04] *** redgone has quit IRC[20:48:06] *** zaph has quit IRC[20:50:58] *** zaph has joined #openindiana[21:04:27] *** GateKeeper has joined #openindiana[21:04:40] *** GateKeeper is now known as Guest98279[21:06:56] *** synegy34 has quit IRC[21:38:52] *** datadigger has quit IRC[21:39:33] *** datadigger has joined #openindiana[22:01:57] *** pasik has quit IRC[22:01:58] <DontKnwMuch> should I have two 1gb bound together or 1 for smb and other for nfs/iscsi traffic?[22:02:17] *** reddi has quit IRC[22:03:47] <kdavy> DontKnwMuch: i'd say bond them[22:04:51] <DontKnwMuch> I have some 3com 4200 switch, more or less old I think, 48 port one.. hope I can figure out how to bond it, I have to setup the switch too, right?[22:05:23] *** pasik has joined #openindiana[22:07:51] <kdavy> yeah i think so. i don't use network for storage[22:13:01] *** DanaG has joined #openindiana[22:13:14] <DanaG> Say, isn't useradd supposed to create a dataset for the user's home dir?[22:13:21] <DanaG> It didn't, when I tried it.[22:14:41] <Triskelios> useradd -m -m, it's disabled by default[22:15:24] <DanaG> ah.[22:17:29] <DanaG> Also weird: when I try to replicate a username that's on my mom's imac, it warns me that the username is too long... yet seems to work somewhat, anyway.[22:19:00] <tomww> olny few programs, mostly 3rd party, might get into troubles.[22:19:39] <tomww> there has been a discussion long ago to enlengthen the official limit for the usernames[22:19:55] <DanaG> could not chdir to (user dir).[22:20:29] <DontKnwMuch> what is the limit for usernames?[22:20:59] <DanaG> That username is 13 characters. I think it's ridiculous, but it seems that's what OS X made by default.[22:21:16] <DanaG> It also didn't create the zfs volume, even with -m -m[22:22:23] <DanaG> Okay, so it looks like -m -m is one thing the "too long" broke![22:22:56] <DanaG> Now it created the dir in /export, but it's not a volume.[22:23:30] <DontKnwMuch> oh...[22:25:24] <DontKnwMuch> if my sistem drive suddenly fails, will I be able to install a new system on new drive and import the pools, exactly as if I made the export before, or will this not work in this way[22:26:06] <geoff_> You can import the data pools.[22:26:29] <geoff_> As long as the OS supports that zpool/zfs version.[22:27:12] <DontKnwMuch> so it is actually not so bad, i just do a network setup, daa the pools back and I am up and running ...[22:27:21] <DontKnwMuch> daa=add[22:28:13] <geoff_> One of the selling points of zfs :)[22:29:48] <DanaG> hmm, skel doesn't contain .bashrc![22:30:25] <DontKnwMuch> :)[22:30:56] <tsoome> smeone should clean that current skel:P[22:31:41] *** pasik has quit IRC[22:32:28] <DontKnwMuch> is there some script, to backup the 'setup' and then to 'restore' it back somehow, like you can do with a router (wishful thinking)[22:32:50] *** sergefonville has joined #openindiana[22:33:20] <DanaG> How do I check whether a user has permissions (through gnome) to initiate shutdown... without actually trying to shut down?[22:34:30] <sergefonville> is there some privilige I can grant so a non priviliged user can create unix sockets?[22:35:12] *** pasik has joined #openindiana[22:35:49] <tsoome> ee user wont have shutdown in menu if he cant[22:37:06] <sergefonville> I'm trying to run uwsgi with a non-privileged user, but I get an error about creating the .sock file[22:37:27] <tsoome> where its attempting to create it?[22:37:39] <sergefonville> is specified /var/run[22:37:54] <sergefonville> the same path I used for php fastcgi[22:38:03] <sergefonville> but tha tis started with spawn-fcgi[22:38:15] <tsoome> and hthat user has permission to create file there?[22:38:15] <DanaG> How do I remove that privilege, then?[22:38:37] <DanaG> Simply removing that permission entirely for everyone, would work too.[22:39:32] <sergefonville> for spawn-fcgi all users run as webservd and the permssions for the .sock are changed accordingly[22:40:04] <sergefonville> with uwsgi I can specify uid and gid, but it switches to that user, prior to creating the socket[22:40:57] <sergefonville> also, there is a message when started without parameters[22:41:06] <sergefonville> that stdin is not a socket[22:41:13] <sergefonville> does that mean it can be?[22:41:55] <tsoome> sure it can[22:42:50] <sergefonville> how do I tell it that is can use stdin as aa socket, that way I could use spawn-fcgi. I think...[22:43:21] <tsoome> that whats happening when you spawn service from inetd...[22:46:18] *** bens1 has joined #openindiana[22:49:48] *** bens1 has quit IRC[22:57:27] <sergefonville> I think it works, at least it starts as it should[22:57:43] <sergefonville> next is to configure nginx to use it[22:57:52] <sergefonville> thanks a lot so far tsoome[22:58:07] <tsoome> didnt do much:P[22:58:34] <sergefonville> and yet, it was enough to be very helpful :P[22:58:37] <tsoome> with sockets you need to make sure the dir where socket is created got permissions for that user[22:58:56] <tsoome> and you can use truss to see whats going on:P[22:59:14] <sergefonville> I used spawn-fcgi like I did with php[22:59:25] <sergefonville> at first that did not work, but now it does :D[23:00:45] <DontKnwMuch> should I make the fs (nfs share for esxi) compression on, or will this make it way slower (compression on for other fs I have, does make some 1.4x but there is no noticable speed difference)?[23:02:21] <kdavy> yes, use compression[23:04:39] <DontKnwMuch> experience?[23:05:11] <richlowe> So, maybe a month ago someone gave me a crashdump that was 60+ frames of nested trap.[23:05:16] <richlowe> ring a bell with anyone?[23:06:55] <kdavy> DontKnwMuch: yes, experience[23:07:21] <richlowe> In general, with lzjb, any speed difference will be positive[23:07:22] <kdavy> 6000+ IOPs and 400MB/s with compression, in production[23:07:44] <richlowe> most modern CPUs can do simple compression/decompression considerably more quickly than most modern storage can move data around.[23:07:45] <kdavy> on single Xeon 5620[23:07:52] <DontKnwMuch> oh. ok. this is good enough for me ;)[23:08:21] <DontKnwMuch> I have E3 1240... but still fast enough I think[23:10:03] <DontKnwMuch> are you using nfs or iscsi too?[23:12:08] <sergefonville> is using a unix-socket bad for performance?[23:14:50] *** jaimef has joined #openindiana[23:15:48] <jaimef> amazing how much load a hung nfs mount generates to the server when the export is no longer available[23:18:37] <DontKnwMuch> ah, one more, should I set the fs blocksize if used for esxi datastores or leave it default, some set it to 128k or somthing, will it make any difference[23:21:27] <tsoome> 128k[23:21:33] *** Shadow__X has joined #openindiana[23:22:35] <Shadow__X> hey guys does openindiana have the zfs administration web frontend thats talked about in this article(http://blogs.oracle.com/talley/entry/manage_zfs_from_your_browser)? or is napp it the one to use[23:23:04] <DontKnwMuch> napp it is ok[23:23:54] <alanc> no, that browser interface used a framework which was not open source, so it was never released in OpenSolaris and was even removed from Solaris 11 because the framework was EOL'ed[23:24:14] <Shadow__X> oh ok that clears that up thanks alanc[23:24:29] <Shadow__X> DontKnwMuch: do you have any other recommendations?[23:24:52] *** skeeziks has joined #openindiana[23:25:13] <DontKnwMuch> didnt even know that the thing you linked exists, does it?[23:25:36] <Shadow__X> well apaprently not anymore but i know napp it is still around[23:25:41] <DontKnwMuch> ah.. missed alanc answer[23:25:44] <Shadow__X> and very easy to install[23:26:00] <DontKnwMuch> nappit helped me a lot[23:26:45] <dkeav> whatever happened to fishworks[23:26:57] <Shadow__X> really? do you mind me asking how it helped?[23:28:12] <Shadow__X> also is there a way to have snapshots automatically created on the even data changes within a directory?[23:28:52] <Shadow__X> the thought there is to be able to have multiple revisions within a document folder, unless there is a better option for document revisions[23:29:06] <tsoome> define data change[23:29:08] <alanc> dkeav: fishworks is now the Oracle ZFS Storage Applicance[23:29:21] <tsoome> if i write 1 byte, is that change enough?:P[23:29:39] <alanc> none of the additional software it includes, like it's browser interfaces for administration & analytics were ever open sourced[23:29:40] <Shadow__X> tsoome: well editing a word document for example[23:29:53] <tsoome> thats bad example[23:30:04] <dkeav> alanc: ahh[23:30:05] <dkeav> bummer[23:30:21] <dkeav> that was an impressive setup[23:30:24] <tsoome> because, if i will write an series of data, but snapshot is taken in between, i can end up with broken file...[23:30:43] <tsoome> (broken on rollback)[23:30:49] <Shadow__X> hmm right[23:31:01] <alanc> though the former fishworks guys who landed at joyent are doing similar stuff with web interfaces to dtrace there too[23:31:34] <tsoome> but then again, there is an timeslider making frequent snapshots, that may be enough...[23:31:41] <alanc> (see bmc & bgregg blogs for examples)[23:31:56] <tsoome> but timeslider soe not know about your application state either[23:32:00] <Shadow__X> tsoome: i have seen the timeslider application but is that limited to the gui only?[23:32:06] <tsoome> no[23:32:34] <tsoome> timeslider has 2 parts - one is snapshotting another is browsing snapshots from gnome file browser[23:32:55] <tsoome> snapshotting does not depend on gui[23:33:27] <Shadow__X> i knew about zfs snapshots but wasnt sure how timeslider was implemented[23:35:19] <tsoome> its all about zfs snapshots + gui for browsing + gui for simple management + addons to make it possible to create backups[23:35:44] <Shadow__X> i gotcha[23:37:40] <Shadow__X> its probably not a good idea for documents then. What i really would like is something like dropbox's implementation of document revisions. I have looked into cvs and subversion but was not sure if there was a better way[23:38:33] <tsoome> well, as if the frequent snapshots are frequent enough, you have good chance to have good one as well. but yep, the most safest way is user initiated snapshots[23:40:06] *** forquare has joined #openindiana[23:41:46] *** forquare has quit IRC[23:44:32] <blues> i'm more excited than i should be[23:44:40] <blues> but I finally got this god damned card flashed[23:44:55] <jkimball4> hmm gnome seems to be randomly remembering sound settings now..which is odd[23:48:44] <blues> http://img839.imageshack.us/img839/7951/59238623424d5b724c00b.th.jpg[23:48:52] <blues> my high-tech flashing station[23:49:21] <dkeav> o.0[23:55:11] <jkimball4> looks kind of small[23:59:37] *** hsp has quit IRC[23:59:42] <DontKnwMuch> another mb? and flash went ok?