October 12, 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 12, 2006 [00:03:12] <Error_404> what in the hell is a gldm_send ? [00:03:32] <quasi> delewis: EAGAIN could be non blocking on a socket iirc [00:04:24] <Error_404> oh, heh... mystery solved... it's a network mutex, and the cable's not plugged in [00:05:06] <Error_404> well, the second mystery anyways [00:05:20] <Error_404> the first... the machine still doesn't boot w/ a sata drive wired up [00:05:35] <quasi> Error_404: short on power? [00:05:52] <Error_404> no, because i can keep the drive powered up [00:06:06] <Error_404> i just can't connect the SATA signalling cable [00:06:07] <quasi> drives use more power on startup [00:06:14] <Error_404> not until after it's POSTed [00:06:50] <Error_404> well in that case i don't understand why it'd be having power issues, it's a 350W PSU powering a sempron, 1 pata drive, and the sata drive [00:07:00] <Error_404> no pci/agp cards even [00:07:37] <Error_404> i'm relatively certain it's a BIOS problem, mostly because i can't think of anything else [00:07:38] <quasi> probably not - it just fit the description [00:07:47] <sickness> i'm back [00:08:18] <Error_404> i'll try yanking power from the cdrom, but i doubt it'll have any effect [00:08:20] <Stric> Error_404: 350W says nothing about the distribution [00:08:30] <Stric> Error_404: or stability of the power [00:10:17] *** dduvall has joined #opensolaris [00:13:10] *** Tpenta has joined #opensolaris [00:13:12] *** B|nTaRa has joined #opensolaris [00:14:47] <Error_404> regardless, it's not the psu [00:18:52] *** Arnald has quit IRC [00:19:10] *** TonyC4 has joined #opensolaris [00:19:43] *** TonyC4 has quit IRC [00:20:46] *** Sieghard has quit IRC [00:21:39] *** Sieghard has joined #opensolaris [00:24:20] <_william_> gn [00:24:21] *** _william_ has left #opensolaris [00:28:02] <Error_404> drats, i have no way of booting the bios setup thingy [00:28:44] *** Disorganized has quit IRC [00:31:04] *** Mazon is now known as mazon [00:35:50] *** metasoup has joined #opensolaris [00:36:05] <Error_404> damn me and my rush to legacy-free! [00:41:06] *** piwi has joined #opensolaris [00:41:22] <boyd> Error_404: heh... I've been there. What about a CD? [00:41:52] <boyd> Ah... after scrolling back I see that may not be an option [00:42:08] <Error_404> yes, i also don't have windows for which to make a dos boot CD [00:42:30] * boyd nods. Bastards. [00:42:54] <dwc-> freedos? [00:42:56] <Error_404> there really should just be a bios that lets you update it from within the bios itself [00:43:03] <Error_404> dwc-: *blink* [00:43:07] <Error_404> i hadn't even thought of that [00:43:13] <dwc-> with a boot image stuffed onto an el-torito bootable cd? [00:43:25] <dwc-> or floppy, if you have one of those things around [00:43:50] <dwc-> some machines, you can netboot a dos floppy image to run bios updates from [00:44:25] *** gdamore has quit IRC [00:46:47] <piwi> hi all, is blogs.sun.com (or maybe the db) "down"? getting only 404 messages [00:47:35] <stevel> seems to be down [00:49:28] <Tpenta> there was a scheduled outage earlier in the day; maybe it ran over time? [00:50:21] <stevel> i'm talking to the WPE guy at the moment anyway, so i let him know [00:50:25] <stevel> he said they'll look into it [00:50:38] <piwi> great [00:53:57] *** metasoup has quit IRC [00:54:34] *** ChanServ sets mode: +o Tpenta [00:54:42] *** Henkis has quit IRC [00:56:18] <dwc-> webserver peon employee? web portal engineer? [00:59:23] *** Cybernd has quit IRC [01:01:03] *** nachox has joined #opensolaris [01:02:25] *** jsubl2 has joined #opensolaris [01:04:44] *** stevel changes topic to "Latest releases: SXCR: 49, ON build: 50, ON nightly: 20061009" [01:14:20] *** nwf has quit IRC [01:19:10] *** bengtf has quit IRC [01:25:54] *** Tpenta has quit IRC [01:27:56] *** Tpenta has joined #opensolaris [01:28:09] *** jmcp_ has joined #opensolaris [01:28:46] *** pseudo_ has joined #opensolaris [01:28:50] *** wesw has joined #opensolaris [01:31:25] *** adp has joined #opensolaris [01:32:11] <coffman> Error_404: a normal "modern" bios can be flashed from bios... (normaly some hot key at startup) [01:32:25] <Error_404> not on my machine [01:32:40] <coffman> amd 64? [01:34:01] *** nachox has quit IRC [01:34:43] <Error_404> coffman: yeah [01:34:56] <Error_404> just a cheap motherboard (obviously, it won't work with this sata drive [01:35:22] <coffman> let me guess no nvidia - via or sis? [01:35:44] *** Sieghard has quit IRC [01:35:49] <Error_404> sis [01:36:06] <coffman> (no one is realy that dumb and buys ati chipsets) [01:36:26] <coffman> u thats shitty [01:36:38] <Error_404> meh, the motherboard was $60, the processor was $70 [01:36:54] <coffman> sata controller in the southbridge? [01:37:40] <Error_404> it would appear so [01:38:08] *** stevel has quit IRC [01:39:18] <coffman> and its not a sil one? [01:39:25] <wesw> forgive me, as i've been away from solaris for awhile, but when is zfs expected for installing on root [production]? [01:39:46] <Error_404> sil? [01:41:01] <coffman> silicon like sil3112 or 3124 [01:42:13] *** deather_ has joined #opensolaris [01:42:15] <Error_404> i don't remember, tbh [01:42:16] *** Xh4 has quit IRC [01:42:20] <Error_404> sis 964 southbridge [01:43:26] <coffman> fake raid ? [01:44:39] *** jcsmith has quit IRC [01:44:40] *** solarisjon has quit IRC [01:49:11] <Error_404> it has the option, but i don't use it [01:49:33] *** dclarke has quit IRC [01:50:10] <coffman> disable it... hm [01:50:44] <coffman> otherwise get a cheap sil3112 or so [01:51:04] <coffman> its better then waisting time with such shit [01:51:50] *** ChanServ sets mode: +o Tpenta [01:53:53] <coffman> well [01:53:56] <coffman> n8 [01:54:06] *** coffman has quit IRC [01:56:47] *** tkup has quit IRC [01:57:05] *** mikaeld has quit IRC [01:59:03] *** deather has quit IRC [02:03:38] *** wesw has quit IRC [02:09:49] *** mikaeld has joined #opensolaris [02:22:06] *** sommerfeld has quit IRC [02:32:48] *** ohtar has joined #opensolaris [02:35:45] *** jsubl2 has quit IRC [02:35:51] *** jsubl2 has joined #opensolaris [02:40:03] *** kimc has joined #opensolaris [02:45:10] *** Peanut has quit IRC [02:48:49] *** Peanut has joined #opensolaris [03:05:57] *** jsubl2 has quit IRC [03:06:35] *** jsubl2 has joined #opensolaris [03:08:51] <dwc-> hm. I just got spam from sun [03:09:18] <dwc-> or wait... no.. [03:09:30] <dwc-> those are fake headers [03:18:00] <LeftWing> Watch out, there was some mob in Singapore that went to the lengths of actually calling me on my mobile to try and scam me into presumably giving up credit card details for the "Try & Buy" programme. =P [03:18:17] <LeftWing> It started with an e-mail with fake headers and then escalated to a phone call. [03:23:21] *** Gman has joined #opensolaris [03:26:27] <Error_404> jesus hell... i'm getting some *terrible* write performance on this drive [03:27:06] <Error_404> zpool iostat's giving me ~ 3M/s [03:27:25] <Error_404> my network's faster than that [03:27:52] <delewis> what's the CPU utilization when it's doing the IO? [03:28:40] <Gman> hey dudes! [03:28:54] *** elflord_ has joined #opensolaris [03:30:39] <Error_404> well, i'm tar.bz-ing a directory [03:30:43] <Error_404> so pretty close to 100% [03:30:56] <delewis> try doing something that *shouldn't* be CPU intensive [03:31:03] <delewis> it sounds like you may have some DMA issues [03:31:11] <Error_404> it's a sata drive [03:31:15] <delewis> how's the performance on UFS? [03:31:50] <Error_404> i don't really remember [03:31:57] <Error_404> the whole drive's been given to zpool [03:32:47] * sparc-kly|WORK is away: go to my house [03:33:53] <delewis> iostat can be helpful :-P [03:34:32] <Error_404> it sure can [03:34:45] <Error_404> regardless, it appears DMA is on, or so says /var/adm/messages [03:34:50] <hile_> hey Gman [03:35:09] <Error_404> meh, i'll just blame bzip2 for this & continue [03:35:49] *** elflord has quit IRC [03:41:38] <elektronkind> buh [03:44:45] *** pseudo_ is now known as Xh4 [03:59:09] * jmcp_ wanders off for lunch [03:59:19] <elektronkind> erm [03:59:27] * elektronkind needs to find someone familiar with libdiskmgt [04:04:57] *** sstallion has joined #opensolaris [04:05:26] <sstallion> evening all [04:05:33] *** Gman has quit IRC [04:06:38] *** piwi has quit IRC [04:07:08] *** Gman has joined #opensolaris [04:09:16] *** jcsmith has joined #opensolaris [04:11:07] *** Kush- has joined #opensolaris [04:15:37] *** nachox has joined #opensolaris [04:16:09] *** Sieghard has joined #opensolaris [04:29:29] *** sstallion has left #opensolaris [04:29:43] *** gdamore has joined #opensolaris [04:29:52] <gdamore> hi * [04:30:13] <elektronkind> HAH [04:30:38] <elektronkind> I'm now several steps closer to getting a reasonably basic Solaris::DiskMgt perl module [04:31:42] <gdamore> I see ON 50 is out now. I wonder, should I bother to download SXCR 49, or should I just wait for SXCR 50? [04:32:06] *** halton has joined #opensolaris [04:32:27] *** janus_man has quit IRC [04:32:54] *** Dr_Jekyl1 has joined #opensolaris [04:33:20] *** janus_man has joined #opensolaris [04:36:44] *** sstallion has joined #opensolaris [04:41:54] *** laca has quit IRC [04:43:14] *** solaris_user has joined #opensolaris [04:45:15] *** Xh4 has quit IRC [04:45:50] *** Xh4 has joined #opensolaris [04:46:21] *** solaris_user has quit IRC [04:46:40] *** Dr_Jekyll has quit IRC [04:47:38] *** dunc_ has quit IRC [04:48:37] *** slowhog has quit IRC [04:49:18] <alanc> we don't even have WOS build 50 internally yet, so SXCR build 50 isn't likely to be out for about a week [04:49:35] *** alanc is now known as alanc-away [04:50:25] *** halton has quit IRC [04:53:57] *** slowhog has joined #opensolaris [04:54:23] *** ohtar has quit IRC [04:55:33] *** slowhog has quit IRC [04:59:38] *** Triskelios has joined #opensolaris [05:18:45] *** sparc-kly has quit IRC [05:25:13] *** Xh4 has quit IRC [05:26:04] *** ohtar has joined #opensolaris [05:30:28] *** J03- has joined #opensolaris [05:33:26] <J03-> wow. quite a crew in here [05:35:36] <J03-> and talkative to boot [05:36:06] *** slowhog has joined #opensolaris [05:39:35] *** slowhog has quit IRC [05:40:19] *** slowhog has joined #opensolaris [05:42:44] *** J03- has left #opensolaris [05:47:16] *** Kush- has quit IRC [05:55:38] <Error_404> i figgured out my drive problem... [05:55:50] <Error_404> apparantly zfs fucks up a drive something awful [05:56:09] <Error_404> solaris has no problem with it, but my BIOS can't make heads or tails of it [05:56:16] <Error_404> so it hangs on POST [06:00:08] *** slowhog has quit IRC [06:00:15] <jamesd> yes its a known bug.... [06:00:30] <jamesd> zfs creates EFI label on drives that it owns. [06:01:11] *** slowhog has joined #opensolaris [06:01:24] <Error_404> i see [06:01:47] <Error_404> that... really kinna sucks [06:02:17] <jamesd> it also can be considered a bug of the bios. [06:02:52] <nachox> the bios should not make assumptios like that [06:02:57] <Error_404> either way, this harddrive is now an expensive brick [06:03:24] <nachox> i'm sure it can be formatted in another box with a less picky bios [06:03:27] <Error_404> unless i get a macpro i guess... they can read efi drives [06:03:30] <jamesd> Error_404, you can move it to another system and wipe the label [06:03:52] <Error_404> meh, it's about time i get another motherboard anyways.... any suggestion? [06:04:12] <Error_404> actually, n/m [06:04:31] <Error_404> will a SATA card solve it, or do you think my bios'd try to read from that too? [06:05:00] <nachox> hmm, sata controllers have their own bioses i think [06:06:44] <Error_404> because that'd be a much more preferable option to buying a new computer [06:06:51] <Error_404> financially anyways [06:08:49] * jmcp_ saunters back in [06:09:23] <Error_404> note that this is a SATA drive, and i can hot plug it back in to the machine [06:09:43] <Error_404> for odd definitions of the term "hot plug" [06:10:05] *** NilsAN has quit IRC [06:10:35] <nachox> hmm, weird, the Java System Directory Server documentation is password protected [06:10:35] *** J03- has joined #opensolaris [06:10:42] <Triskelios> yeah just get a sane controller [06:10:55] *** NilsAN has joined #opensolaris [06:11:24] <Error_404> sil something something (x*12 for some X in N) [06:11:56] <J03-> hey guys i'm very interested in tinkering with ZFS and namely RAID-Z arrays. one of my main requirements is that it support Online Capacity Expansion. there is no way to do this in software is there? [06:12:37] <jamesd> you can add more raidz groups to a pool... [06:12:50] <elektronkind> Error_404: something with a sil3124 or 3132 chip [06:13:03] <Error_404> *nod* [06:13:15] <J03-> right, but that doubles the parity data [06:13:24] <jamesd> you can add 1000's of raidz groups to a pool... having more than 7 or 8 disks in a raidz pool is a bad idea.. [06:13:31] <elektronkind> 3132 is PCIe->SATA2 [06:13:32] <J03-> yeah i understand that [06:13:36] <jamesd> er raidz group [06:13:47] <J03-> i'm wanting to start out with 3 drives and up it occasionally. i dont think there is a way to do this with any type of software raid [06:14:10] <J03-> i probably wouldnt go over 6 drives without creating a new raidz group [06:14:30] <J03-> but adding a minimum of 3 drives at a time is no fun, hehe [06:14:31] <jamesd> J03-, linux says it can do it in raid5... but it has other limits.. including slow writes and no data verification. [06:14:47] <J03-> yeah with EVMS? [06:15:11] <elektronkind> well, if you start out with three drives, make a raidz of them with your new pool, and later when you add 3 more, make those 3 a new raidz set and add that to your existing pool [06:15:14] <jamesd> i guess.. i haven't checked into it.. i researched once.. and promptly dismissed it [06:15:25] <J03-> i've been looking around for a few hours trying to find out if there's anything that fits my needs. so far it looks like linux and EVMS may be my only option, but i dont feel comfortable with software raid unless it's ZFS [06:15:57] <J03-> so i'll likely have to buy a hardware raid card with OCE [06:16:19] * elektronkind doesn't understand what the drawback is here [06:16:20] <J03-> that's a couple hundred bucks wasted in my book, because unless you go with the mega expensive cards, you get very bad write speeds [06:16:32] <jamesd> J03-, besides if you wait more than a year between upgrades.. it would be pretty hard to find drives that are the same size as that you had started with [06:16:49] <J03-> elektronkind, because if you have 3 drives in raidz, when you fill the drives you end up basically losing 1 drive to parity information [06:16:59] <J03-> add another 3 drives to the pool and fill it up and you now have 2 drives full of parity [06:17:16] <J03-> if you had 6 drives to start with and in a single raidz group, after you fill it up you only lose 1 drive to parity [06:17:18] <jamesd> but your data is safe if upto 2 drives fail [06:17:33] <jamesd> and if you do that and 2 drives fail.. your data is gone [06:17:35] <elektronkind> J03-: look at it this way... two raidz groups of 3 drives each is more reliable than 1 raidz group of 6 drives... [06:17:44] <J03-> yeah i know it's more reliable [06:18:09] <J03-> but my point is i dont want to add 3 drives at a time, and i think 2 drives for parity out of 6 drives total is very much overkill [06:18:30] <jamesd> is your data worth more than $200 [06:18:52] <J03-> of course it is [06:19:16] <J03-> but there's no need in wasting $200 when you dont have to or dont find it neccessary [06:19:19] * elektronkind shrugs [06:19:24] <jamesd> then why not spend a litttle extra and guarantee its safe. [06:19:36] <J03-> i could spend $250 once and get a hardware raid card with 8 SATA ports that supports OCE [06:19:48] <elektronkind> have fun trying to grow a linux raid5 set [06:19:56] <J03-> doesnt have to be linux [06:20:04] <J03-> and the hardware card will do it blindly to linux [06:20:11] <J03-> blind to any OS [06:20:22] <jamesd> J03-, and what happens when your hardware raid card dies? [06:20:33] <J03-> same thing when a hard drive dies. i put another one in there [06:20:46] <elektronkind> you'll probably end up spending more on a hw raid card that grows raid5 LUNs than you would on that "wasted" second partiy disk [06:20:57] <J03-> like i said, i would LOVE to do this array in ZFS so i dont have to mess with hardware raid. but OCE is pretty much a must [06:21:00] <jamesd> good luck... they may change the bios chip on it making it incompatilble with the old data [06:21:10] <elektronkind> expandable hw raid5 aren't exactly consumer-level items [06:21:28] <J03-> sure they are [06:21:35] <nachox> no, the cards usually have some sort of memort that stores the state of the disks and the like, it's not as easy as changing the card [06:21:58] <elektronkind> I meant expandable as in "will grow a raid5 LUN" [06:22:09] <J03-> they have NVRAM on them, but it gets cleared every time you have a full stripe to write [06:22:38] <J03-> so i guess the answer to my original question is "no"? [06:22:57] <elektronkind> it also means that if your hw RAID5 card dies, so does your data, unless there's a mechanism to back up the card's config. [06:23:04] <elektronkind> zfs doesn't suffer from that [06:23:47] <J03-> i'm aware of the positives and negatives of the different scenarios [06:23:57] <J03-> it's not truly backed up unless you have an off site copy anyway [06:24:11] <J03-> i want an array that i can grow without wasting too much space in parity [06:24:13] <elektronkind> and that's another subject altogether [06:24:39] <J03-> i just figured ZFS may have the feature i'm looking for, just not very well documented or something [06:24:42] <J03-> or maybe it's in the works? [06:25:06] <J03-> i know it can be done with linux or with hardware raid. i was hoping ZFS allowed it [06:25:08] <elektronkind> zfs takes the view that disk is cheap [06:25:48] <nachox> what do you mean with the parity wasting too much space? [06:26:14] <J03-> two seperate raidz arrays use more parity than a single array does [06:26:31] <jamesd> J03-, buy cheaper smaller disks... enable compression if your data is suitable and most likely get more usable space than most OS's can achieve. [06:26:44] <elektronkind> nachox: he's referring to the lost usable space one gets when using 2x three drive raidz groups in a pool vs. 1x six drive raidz group [06:26:59] <jamesd> for less than a smaller number of larger drives. [06:27:12] <nachox> elektronkind: makes sense [06:27:43] <J03-> jamesd, i appreciate the suggestions, really. [06:27:49] <J03-> but i want OCE [06:27:58] <elektronkind> wtf is OCE [06:28:03] <J03-> if it cant be done with ZFS, so be it. is it ever planned? [06:28:10] <J03-> online capacity expansion [06:28:39] <J03-> Online Capacity Expansion and Online RAID Level Migration (OCE/ORLM) [06:28:50] <elektronkind> if you're comfortable with playing tricks like that with your data, go for it. [06:29:21] <J03-> http://www.newegg.com/Product/Product.asp?Item=N82E16816115026 [06:29:34] <J03-> there's a cheap "hardware" raid5 card that will do it [06:29:39] <elektronkind> seems like a very small thing to worry about if your data is that important. [06:29:45] <J03-> technically it's a software card since there's no nvram or xor processor, but whatever [06:29:58] <elektronkind> J03-: there's a diff between "will do it" and "will do it well" [06:30:03] <J03-> it would accomplish my goal. only problem is i dont want ot waste money on a raid card. also it has slow write speeds [06:30:12] <jamesd> i have a friend that had to deliver the news more than once that because some company decided to save the cost of an single additional drive the company now has to be $50,000 to restore the data because one more drive than they had planned for failed before it could be fixed. [06:30:50] <J03-> ok fine, my data is NOT that important. whatever [06:30:51] <nachox> jamesd: life's a bitch :P [06:30:52] <elektronkind> J03-: oh for crying out loud, either way, you're going to "waste" money on SOMETHING. welcome to the world of having to pay to play [06:31:05] <J03-> i wouldnt if ZFS supported it [06:31:06] <elektronkind> more disk, or raid card [06:31:11] <J03-> which is why i came here to ask [06:31:12] <elektronkind> cake <=> eat [06:31:58] <nachox> you can add more disks to the pool to increase the capacity and i think it can be done online but i'm not sure [06:32:20] <jamesd> yes it can be done online. [06:32:21] <J03-> yeah you can add to the pool, but not to the raidz group [06:32:40] <J03-> am i correct? [06:32:49] <jamesd> yes you are correct... [06:32:50] <nachox> and you can get a pool of disks and for example mirror it but i dont know if you can get a mirrored pool and "downgrade it" [06:32:58] <J03-> i'm no solaris guru. i just installed it for the first time last week and i've been soaking up the ZFS goodness [06:33:09] <J03-> it's just missing that ONE feature that would make it absolutely perfect for me [06:33:11] *** nwf has joined #opensolaris [06:33:12] <jamesd> but try and buy a 250GB drive in 2-3 years... what is common today won't be in 2 short years. [06:33:29] <J03-> you'd be surprised man [06:33:40] <J03-> i've had my current hardware ATA raid5 array going for 3 years [06:33:43] <jamesd> you will be buying a used drive... [06:33:55] <J03-> 4x250gb drives. when one died a few months back i was able to buy an exact match [06:33:59] <jamesd> J03-, and can you buy more drives of that size? [06:34:08] <J03-> yeah [06:34:15] <J03-> i can still buy the same drives i used to start with [06:34:31] <J03-> even so, you can always use slightly larger drives [06:34:43] <J03-> it's optimal that they match, but not a must [06:34:46] <jamesd> but you would be waisting space. [06:34:57] <nachox> wasting the space you now want to spare in parity [06:34:59] <J03-> yeah but not wasting an entire drive [06:35:23] <J03-> and i'm not talking about 2-3 years from now. i'm talking every few months i may want to add a drive [06:35:31] <J03-> as i save money, etc [06:36:17] <jamesd> just use your current solution for 6 more months.. and continue to save. [06:36:54] <J03-> current solution is pretty much full [06:37:09] <J03-> and if the card i have now supported OCE, i could easily add disks [06:37:11] <elektronkind> if your data is easily compressable, a 2:1 compression ratio in zfs will turn, say, a 500GB raidz into an effective 1TB raidz set [06:37:21] <J03-> it's not compressable [06:37:31] *** Netwolf_ has joined #opensolaris [06:37:56] <J03-> by the way how cpu intensive is compression on ZFS? [06:38:00] <J03-> i havent tinkered with it [06:38:20] <nachox> not very [06:38:41] <Netwolf_> J03-: I've heard varying reports that zfs os a bit cpu intensive [06:38:46] <elektronkind> not much. these days stuff like that isn't cpu-bound. I actually get higher IO rates in some cases with compression turned on since it's reading/writing less data to the disks [06:38:47] <J03-> that would probably be very useful for databases [06:39:22] <J03-> probably depends on the number of drives, connection throughput, and the cpu power [06:39:38] <nachox> there are some blogs about zfs and databases, you would do well reading them [06:39:56] <J03-> i've read a ton of them [06:40:29] <J03-> i'm not VERY linux/unix savvy, so i cant delve into the ambiguarities of each very much [06:40:35] <J03-> but i know enough to get myself into trouble [06:40:55] <J03-> i have build 48 running in vmware playing with the ZFS pool stuff [06:41:04] <elektronkind> zfs compression uses lzjb, which is a (IIRC) optimized lz compression algorithm [06:41:06] <J03-> cant really run reliable benchmarks inside of a virtual machine [06:41:25] <J03-> trying to explore my options before i decide to build the box [06:42:03] <J03-> i guess one option would be to play with the EVMS and linux with the different file systems [06:42:12] <J03-> my one fear is the raid5 write hole [06:42:27] <J03-> i dont think there's any type of linux filesystem that would correct such [06:43:13] <jmcp_> J03-: what raid5 write hole are you talking about re zfs? [06:43:16] <nachox> sure there is, reiserfs, hans hid his ex there ;) [06:43:51] <J03-> there is no write hole with ZFS [06:44:18] <J03-> the write hole comes from writing partial stripes to a disk [06:44:41] <J03-> if the power goes out or something you could possibly be boned and never know it untill you tried to access certain data [06:44:44] <J03-> which would come back corrupted [06:45:30] <J03-> ZFS corrects it by treating blocks as an individual stripe and CRC'ing the snot out of everything [06:45:43] <J03-> fixing anomalies [06:46:09] <jamesd> i would be scared of a lot more than simply write holes... 2 drive failures would give me fear as well and if you are using linux, a new kernl version and defective driver would also cause me some fear as well. [06:46:16] <nachox> looks like he's been reading jeff's blog, hehe [06:46:53] <J03-> haha, probably. i'm a reader [06:46:56] <nachox> jamesd: the defective driver problem is a general problem with software, it's not linux alone [06:47:15] <J03-> the 2 drive failure thing doesnt really concern me. it's easy to add a hot swap [06:47:20] <J03-> err, what's it called [06:47:34] <J03-> makes it similar to raid6 anyway [06:47:38] <jamesd> nachox, true... but i would trust solaris is a bit more stable in drivers [06:48:10] <nachox> only when acpi is not involved ;) [06:48:12] <J03-> i'm very impressed with solaris so far. i havent ran it in years since i used to run a Tribes server on a SPARC II i think it was [06:48:17] <jamesd> J03-, it 2 drives fail before you get a replacement and it the array is scrubed/resilvered... its still dead. [06:48:19] <J03-> college days [06:48:27] <J03-> hot spare, hah [06:48:29] <J03-> that's what it's called [06:48:41] <J03-> cant you enable a hot spare on the array with ZFS? [06:48:54] <J03-> if one drive dies, it takes over for that drive automatically [06:49:03] <J03-> thereby keeping parity intact while you replace the failed one [06:49:20] <jamesd> yes you can... but now you are waisting drive space anyway.. so why not just use an extra drive and have a pair of raidz pools [06:49:39] <J03-> because as it grows the wasted parity is exponential [06:49:49] <J03-> adding a hot spare i at most will lose 2 drives [06:49:58] *** Netwolf has quit IRC [06:50:08] <J03-> which i think would be sufficient for a 12 disk array [06:50:19] <J03-> if i created four 3 disk arrays, there's 4 drives wasted [06:50:24] <nachox> heh, harddrives are so inexpensive nowadays i dont even know why we're arguing [06:50:51] <J03-> hard drives ARE relatively expensive, but getting them hooked up in an enclosure of some sort isnt [06:51:54] <J03-> http://www.extremetech.com/article2/0,1558,2018300,00.asp [06:52:11] <J03-> my goal would be to buy that enclosure and run it in JBOD mode on the eSATA controller [06:52:20] <J03-> yet i'd let ZFS control the raid sets and stuff [06:52:29] <nachox> search for thumper :P [06:52:42] <J03-> i'd be able to grow the array up to 12 devices in this scenario [06:52:47] <J03-> thumper? [06:53:20] <jamesd> x4500] [06:53:22] <J03-> my ideal situtation would be to start out with 3 drives, and add them occasionally as i need space, or if i think it neccessitates a hot spare [06:53:49] <J03-> what? [06:54:21] <J03-> oh pshhhh, $32k [06:54:24] <J03-> where's my amex [06:54:41] <jamesd> http://store.sun.com/CMTemplate/CEServlet?process=SunStore&cmdViewProduct_CP&catid=151017 [06:55:14] *** karrotx has joined #opensolaris [06:55:25] *** karrotx has joined #opensolaris [06:55:26] * J03- picks his jaw up off the floor [06:55:56] <nachox> i'm off to bed, i have to work tomorrow... i'll see if i can get my coworkers to join me when going to the tech days in buenos aires :) [06:55:59] *** dunc_ has joined #opensolaris [06:56:01] <J03-> i could outsource my data backup to india for that price [06:56:23] <nachox> and what do you think they will use in india to store it? ;) [06:56:24] <J03-> ie - have a bunch of indians constantly burning dvd-rs for me [06:56:42] <J03-> probably haxxed ipods or something [06:56:52] *** nachox has quit IRC [06:56:55] <boyd> Good luck with that. [06:57:00] <J03-> anyway, very interesting discussions [06:57:17] <J03-> but, any clue as to if they plan on adding this feature to ZFS? [06:57:36] * boyd just arrived... what feature [06:57:37] <boyd> ? [06:58:13] <J03-> ...you cant scroll up? [06:58:14] <J03-> it's quite a few pages [06:58:28] <J03-> namely Online Capacity Expansion if you know what that is [06:58:33] <boyd> If you can't be bothered typing it I can't be bothered answering. [06:58:46] <J03-> i already did type it :o) [06:59:12] <J03-> i'm fairly certain it's not a feature now. basically wondering if it's on the drawing board [06:59:32] <boyd> I know what it is. It's on the list. Check the zfs-discuss archives for more info, but I'm pretty sure it's no more than on the list ATM. [07:00:09] <boyd> Of course, you no doubt know that you can do it at the pool level. [07:00:10] <J03-> nice to know it's considered [07:00:20] <J03-> yeah but uses 2x the parity [07:00:28] <J03-> and you have to add disks 3 at a time [07:00:47] <boyd> Actually, you can add 2 [07:00:53] <J03-> with OCE, i have many more versatile options [07:00:59] <J03-> well, yeah [07:01:37] <J03-> OCE seems like the last link in the chain to allow ZFS to do just whatever the hell you want with it [07:01:53] <J03-> with never having to destroy an array [07:02:23] <boyd> It certainly would be nice, but a little bit of planning can reduce the impact of that limitation. [07:02:33] <J03-> yeah i understand that [07:02:38] <jmcp_> "double the parity, double the fun" is what the raidz-2 RFE justification reads [07:02:41] <J03-> by my plan is to add single drives incrementally, heh [07:02:51] <J03-> *by/but [07:02:59] <boyd> Hey, jmcp [07:03:05] <boyd> err jmcp_ :) [07:03:11] * boyd greeted the wrong clone :) [07:04:28] <J03-> can you help me out with a link on the ZFS discussions? preferably one to the OCE specifically [07:05:10] <J03-> also, who should a prod? heh. i could possibly build my array now and maybe the feature will be there by the time i need to add disks? [07:05:47] <J03-> but, i havent delved much into upgrading solaris, no clue how deep that rabbit hole would go [07:06:15] * boyd looks for a URL, since he normally reads via email [07:09:23] <boyd> Aw this is the wors search engine in the world. Oh no, sunsolve is worse [07:09:42] <boyd> I have the subject line but I can't find the thread [07:09:51] <twincest> i suspect b.o.o might easily come second, too [07:10:09] <boyd> True [07:11:13] <boyd> Ok, so this thread seems like the most recent occurrence: http://mail.opensolaris.org/pipermail/zfs-discuss/2006-August/004386.html (Thanks google) [07:11:59] <J03-> hah, i've already read that before [07:12:07] <J03-> i dont see a response though [07:12:20] <J03-> basically that's all i can find. people asking questions but never an official response [07:12:53] <J03-> or maybe he is answering a question? it's hard to tell because of the bad english [07:14:42] <J03-> oh wait, there it is [07:15:27] <boyd> Sorry, I only found the q, I didn't look for a response in that thread. [07:15:48] <J03-> well they basically responded by saying ZFS doesnt do it right now [07:16:00] <J03-> then they start talkign about Net Filers, whatever that is [07:16:06] <J03-> i'm assuming some type of vendor solution [07:16:14] <boyd> It's hard to tell since the freakin' web forum software breaks threads. [07:16:19] <boyd> Lemme find another one. [07:16:26] <boyd> NetApp Filer == NAS [07:16:30] <J03-> i've already read all this before. it's helpful in letting me know that the feature isnt there now [07:16:36] <J03-> yeah that's what i assumed [07:17:01] <J03-> That's the default, I think, but you can use 'vol add -g' to add disks to an existing RAID group. This is fairly new functionality (V6.2 I think). ZFS will probably not take so long to add this feature. :-) [07:17:01] <J03-> [07:17:18] <J03-> he's referring to NetApp Filer, but then mentions it probably wont take long [07:17:27] <J03-> wonder how official the source is [07:17:40] <J03-> Anton Rang? [07:22:17] <boyd> Oh, wait... I must have been thinking about something else (vdev removal): [07:22:18] <boyd> http://mail.opensolaris.org/pipermail/zfs-discuss/2005-November/000116.html [07:22:43] *** jsubl2 has quit IRC [07:23:05] <boyd> Eric Schork is pretty authoritative on ZFS. [07:23:10] <boyd> Schrock :) [07:25:28] *** karrotx has quit IRC [07:25:49] <J03-> meh [07:26:01] <J03-> he basically said what i suspected [07:26:14] <J03-> and gives other options that dont really work for me [07:26:25] <J03-> although that was almost a year ago [07:27:12] *** Netwolf_ has quit IRC [07:28:41] *** elflord_ has quit IRC [07:29:40] <boyd> Frankly for the price of disk these days I'd rather do raid 0+1 if I wanted incremental expansion [07:30:27] <J03-> my main contraints are money and a place to put the drives [07:30:44] <J03-> the parity is important, but i definately dont forsee 2 drives failing at the same time [07:30:53] <J03-> and if they do, no HUGE loss [07:31:07] <J03-> data management is all about risk/price accessment [07:31:52] <boyd> Wel if those are you main constraints then make raidz sets. [07:32:23] <boyd> You didn't mention resizing in them, so it's not in the top three it seems. [07:33:04] <J03-> well when you waste tons of space on parity, there's added costs. also when you're wasting disks to unneeded/wanted parity, you run out of hot swap bays to put drives [07:33:14] <J03-> which is my reasoning for wanting expansion [07:33:19] <boyd> Hence, raidz. [07:33:26] <boyd> minimal parity [07:33:34] <J03-> exactly [07:33:57] <J03-> maximum storage, minimal parity, and best cost/risk ratio for my needs [07:34:08] <J03-> my only problem is if i want to add single drives to the array, i cant do it [07:34:09] <boyd> I don't see expansion in there anywhere [07:34:10] *** lisppaste3 has quit IRC [07:34:34] <J03-> if i have two 3 disk raidz sets, i have 2 drives lost to parity [07:34:51] <J03-> that's not optimal because i waste money on drives and rack space for the drives [07:34:54] <boyd> So, don't make two sets [07:35:15] <J03-> exactly. i want a larger set. [07:35:16] <boyd> If you consider redundancy to be waste then we're haviing a different conversation [07:35:25] <boyd> So make a larger set then. [07:35:32] <J03-> money [07:35:38] <J03-> easier to start small and grow it [07:36:01] <J03-> and parity isnt a waste, but there's a point wher eit's exorbant [07:36:08] <boyd> Ah, now that's a different problem. [07:36:14] <J03-> right [07:36:33] <J03-> ideally i'd like to start a raidz array with 3 drives and add them ONE AT A TIME as needed [07:36:41] <J03-> currently it's not possible unless you use a hardware raid card [07:36:56] <boyd> I get what you wan't. You can't have it with zfs atm. [07:37:03] <boyd> This conversation is circular [07:37:06] <J03-> but there inlies more money spent and more limited capacity due to the number of ports [07:37:13] <J03-> yeah exactly [07:37:16] <J03-> i've known this from the start, heh [07:37:34] <J03-> my only question when i entered the room is, "is it possible, and will it be?" [07:37:52] <boyd> No. Probably not. [07:37:57] <J03-> it turned into 10 pages of people offering suggestions and weighing the pros/cons [07:38:03] <J03-> appreciated nonttheless [07:38:17] <J03-> yeah that's what i figured [07:38:37] <J03-> so NOW it's time to explore other options, or just cut my losses and spend money that i dont want to [07:39:32] <J03-> http://mail.opensolaris.org/pipermail/zfs-discuss/2005-November/000116.html [07:39:41] <boyd> With veritas volume manager you can resize... but check on space constraints and co$$$$t. There is a free, limited version [07:39:42] <elektronkind> dear lord. my little 133 line C program has 22 libraries linked into it L/ [07:39:47] <J03-> that link, right after where he talks about scenario A and B [07:40:13] <J03-> in the parenthesis he mentions that it only takes a small space penatly [07:40:26] <J03-> i wouldnt call 25% a small penalty [07:40:48] * boyd shrugs. [07:40:49] <J03-> basically 1 out of the 4 drives is doing something that doesnt fit what the user wants/needs [07:41:17] <boyd> No, we assume "not losing data in the case of hardware failure" is part of what the user wants. [07:41:31] *** jengelh has quit IRC [07:41:34] <J03-> then why not force raidz? [07:41:48] <elektronkind> dude, don't be silly [07:41:54] <boyd> As long as you think of parity as wasted space you will always begrudge any of it [07:41:55] <elektronkind> you're making zero sense [07:42:14] <J03-> i dont think of it as wasted. i think of that scenario as excess [07:43:36] <elektronkind> like I said before, zfs take the view that disk is cheap. An extra $200 drive is worth the data retention to a lot of people, and it's not as bad as mirroring [07:44:07] <boyd> Ok, that's your position. It's not the same as that of most people who are interested in keeping their data availiable. [07:44:21] <boyd> ^^ J03- [07:44:27] <elektronkind> yes, you can say "well XX thingy lets me resize" but if you believe that playing such magic tricks with your only copy of precious data is ok, then by all means, go for it [07:44:34] <J03-> you dont think that 1/4 of an entire array is overkill for a home user? [07:45:03] <jmcp_> not at all [07:45:11] <galt> J03-, depends, how much pr0n do you need? ;P [07:45:14] <boyd> Listen, only you can decide that. if it's too much, make a larger set or use no parity. it's your data [07:45:17] <J03-> it's currently what i'm doing. ihave 4x250gb. i've had 1 drive die in 3 years [07:45:41] <J03-> boyd, yeah i know man. it's all down to user preference [07:45:49] <jmcp_> boyd: I'm mirroring the drives in my ultra20m2 which is arriving tomorrow [07:46:00] <J03-> but being able to grow a raidz group seems like a very useful feature to me and gives the customer/user a much more diversified set of options [07:46:30] <boyd> Mu last box I intended to mirror... never did get around to it and now I have other data on there :) [07:47:02] <boyd> J03-: Of course it would be nice. So would world peace. [07:47:13] <jmcp_> boyd: I purchased 2x320gb drives a few weeks ago waiting for the new system [07:47:14] <elektronkind> so would parity-less raid5 [07:47:15] <boyd> It's time to move on. [07:47:39] <boyd> jmcp_: Dude... unless you're more disciplined than me, mirror them the second they go in the box :) [07:48:10] <jmcp_> that's why I didn't put them into the existing boxes :) [07:48:24] <boyd> Heh... welll done... resist the urge. [07:48:25] <J03-> well, in the same vein as world peace, just because it's very difficult and would probably never happen, doesnt mean it shouldnt be attempted and hoped for :o) [07:48:26] <jmcp_> gonna do the UFS root/boot/swap mirror thing, and ZFS for the rest [07:48:33] <elektronkind> jmcp_: you haven't had any experience with libdiskmgt have you? [07:48:43] <jmcp_> nope [07:48:47] <elektronkind> dang [07:48:57] *** Sparcioni has joined #opensolaris [07:48:58] <jmcp_> elektronkind: I did have a look through the code to try to answer your question but I didn't get a chance to follow up with any useful info [07:49:05] <J03-> many many users looking for the same feature as i am, but i digress. mooooooving on [07:49:09] <boyd> J03-: Sure, "if you think you can develop a tool to do this, let us [07:49:09] <boyd> know. The source code is out there for a reason" [07:49:10] <J03-> thanks for the comments/suggestions [07:49:18] <J03-> haha, yeah, i dont think i can [07:49:46] *** elflord has joined #opensolaris [07:50:08] <boyd> jmcp_: Yeah, that's what I'm doing at the moment. With 2xRoot for live upgrade [07:50:35] * boyd made his first branded zone yesterday. Now if only I had something to run in it :) [07:50:56] <J03-> man that zone stuff is confusing, but sounds extremely useful [07:51:02] <J03-> havent had time to play with it [07:51:09] <jmcp_> I was trying to get the linux gropewise client running my branded zone, but couldn't due to nyetworking [07:51:19] <boyd> jmcp_: ? [07:51:38] <boyd> (was that a deliberate typo? it's good if not) [07:51:41] <twincest> boyd: i used mine to run opera, but then they ported it to x86. bastards. [07:52:15] <jmcp_> boyd: with different network ranges the nat stuff wasn't working at all [07:52:26] <jmcp_> and yes it was a deliberate "nyet" [07:53:18] *** silly_girl22 has joined #opensolaris [07:53:37] <silly_girl22> pkg-get wont install mysql5 [07:53:59] <silly_girl22> can anyone help? [07:54:10] <elektronkind> this is what I'm working on: http://pastebin.ca/198829 [07:54:33] <elektronkind> probing the deep ins-and-outs of disks, slices, busses and controllers [07:54:55] <elektronkind> eventually, there will be a libdiskmgt perl module [07:55:27] <boyd> elektronkind: Sweet [07:55:48] <boyd> silly_girl22: Any chance you can provide more info than "wont install"? [07:55:52] <silly_girl22> gzip: mysql5-5.0.24,REV=2006.09.02_rev=a-SunOS5.8-i386-CSW.pkg.gz: unexpected end of file [07:56:00] *** Xh4 has joined #opensolaris [07:56:11] <boyd> sounds like the download broke to me. [07:56:23] <silly_girl22> so how can i get it to re-download? [07:56:39] <boyd> I thought it tried again every time. [07:56:49] <silly_girl22> everytime i try, i get the same error [07:56:55] <silly_girl22> maybe you are right [07:56:58] <boyd> It could be broken upstream [07:57:13] <boyd> I can't check right now... [07:57:26] <boyd> you may be able to ask in #blastwave [07:57:57] *** solaris_user has joined #opensolaris [07:58:29] <solaris_user> anyone has Solaris internals 2nd pdf version? [07:58:49] <boyd> I think the authors do. [07:58:52] <elektronkind> safari.com does [07:59:04] <solaris_user> safari.com does not accept paypal [07:59:06] <solaris_user> 0)0 [07:59:19] <solaris_user> i live in brazil 0)0 [07:59:32] <solaris_user> 8 weeks to send this book 0)0 [07:59:38] <solaris_user> so I need fast :( [07:59:48] <boyd> Geez, what do they use pigeons?! [08:00:01] <silly_girl22> im getting so tired, im seeind double [08:00:18] <boyd> Only took 6 weeks to me in Australia and it went via Germany [08:00:46] <solaris_user> boyd : whats pigeons means??? [08:01:20] *** NilsAN has quit IRC [08:01:34] <boyd> solaris_user: http://images.google.com/images?q=pigeons [08:01:51] <jmcp_> solaris_user: he is asking whether the vendor is using pigeons to carry the book to you [08:01:57] <solaris_user> lol [08:02:02] *** NilsAN has joined #opensolaris [08:02:05] <solaris_user> :) [08:02:11] *** ohtar has quit IRC [08:02:21] <Sparcioni> I am debugging a process on solaris.After some time I see by prstat that the RSS grows. pmap pid | grep heap => shows heap size of a process [08:02:26] <solaris_user> its serious i need that book :( [08:02:29] <jmcp_> hence "carrier pigeon" [08:02:37] <Sparcioni> is there some command/util to monitor a process heap except this "pmap pid ? " [08:02:55] <jmcp_> Sparcioni: does the process' heap go to a certain point and then stabilise, or does it keep growing? [08:03:03] <solaris_user> proc tools are the best way to do this [08:03:09] <jmcp_> and no, I don't know of any other way of monitoring the heap size [08:03:12] <Sparcioni> keep growing [08:03:20] <silly_girl22> can someone try pkg-get -i mysql5 to see if its working? [08:03:23] <silly_girl22> its not working for me [08:03:26] <silly_girl22> please [08:03:46] <Sparcioni> solaris_user, proc tools : what is it ? sorry for my ignorance [08:04:03] <solaris_user> proc tools all p***** programs [08:04:16] <solaris_user> pmap is a prc tool [08:04:20] <boyd> What's wrong with pmap? [08:04:21] <Sparcioni> ok [08:04:22] <jmcp_> Sparcioni: man proc and man pmap [08:04:32] <Sparcioni> jmcp, ok, thnks [08:04:35] <twincest> you could dtrace calls to brk() [08:04:40] <jmcp_> Sparcioni: since the heap keeps growing, you should look at the functions which are calling malloc() and not free() [08:04:49] <jmcp_> twincest: damn, one step ahead! [08:04:51] <jmcp_> :) [08:05:00] <Sparcioni> it in fact in Java [08:05:07] <solaris_user> you can play with dtrace a little [08:05:08] <boyd> Ah. [08:05:18] <solaris_user> u will have the best debug [08:05:22] <boyd> Well Java does tend to do that untill the gc runs [08:05:33] <Sparcioni> but there is NO OutOfMemoryError from JAVA [08:05:43] <jmcp_> Sparcioni: use the DTrace java provider [08:06:12] <boyd> I'm sure there are other java mem tools too. There is some kind of verbose GC info.... it's in SPAT [08:06:22] <Sparcioni> DTracing calls to what ? [08:06:25] <Sparcioni> brk ()? [08:07:03] <boyd> There is a dtrace provide for java... I think it can show java method calls. [08:07:04] <Sparcioni> boyd, there are some profiler from boreland I once used [08:07:14] <solaris_user> u can also play a little with memtool ( but i really dont know if works on solaris 10/opensolaris ) [08:07:29] <jmcp_> Sparcioni: something like this: dtrace -n'java:::' -p pidOfSuspectProcess [08:07:43] <jmcp_> as a starting point, I hasten to add [08:07:55] <Sparcioni> jmcp, ok . thnks ; I will try that [08:07:58] <solaris_user> truss 0)0 [08:08:35] * boyd is trying out google reader again thanks to jmcp_ [08:08:55] <silly_girl22> Installation of <CSWmysql5> partially failed. [08:09:02] <solaris_user> see log [08:10:52] <silly_girl22> 1:12am up 25 day(s), 15:51, 1 user, load average: 0.12, 0.09, 0.07 [08:11:59] <boyd> jmcp_: It misses space going to the next feed at the end [08:12:33] <jmcp_> huh? not sure I understand you [08:12:55] <jmcp_> silly_girl22: why do you think the load average is going to be relevant to your pkg installation problem? [08:13:24] <boyd> jmcp_: I mean that I wish space-bar would go to the next feed if this one is all read [08:13:41] <silly_girl22> i just think its good that i have 25 days without a reboot but i know i can do better [08:13:58] <jmcp_> boyd: got it, right [08:14:05] <twincest> cuz number of days without a reboot is really relevant to... wait, nothing at all [08:14:16] <silly_girl22> who said it was [08:14:19] <solaris_user> silly_girl22 : my mail server has 2 yrs uptime :) [08:14:23] <twincest> can someone please remove any way to print uptime in s11? :) [08:14:26] <silly_girl22> relavent to my problem [08:14:32] <solaris_user> solaris 5.1 :) [08:14:42] *** tsoome has quit IRC [08:14:46] <silly_girl22> solaris_user: can u put the uptime output here [08:14:50] <jmcp_> silly_girl22: I had 45 days on my SX box earlier, then I needed to make some hw changes [08:14:57] <solaris_user> not now just login insite NOC [08:14:58] <twincest> solaris: 2.5.1? [08:15:11] <solaris_user> yeap sparc 20 solaris 5.1 [08:15:14] <silly_girl22> jmcp_: you should have hotswapped. you dont have to turn the computer off. [08:15:17] <twincest> there isn't a 5.1 [08:15:20] <Triskelios> "solaris 5.1" doesn't exist [08:15:24] <solaris_user> 2.5.1 [08:15:27] <solaris_user> same -_- [08:15:42] <twincest> really not the same.. 5.1 could also mean 2.1 [08:15:43] <solaris_user> i relly need put 2?!??!? [08:15:44] <jmcp_> Triskelios: technically it did .... [08:15:46] <solaris_user> lol [08:16:59] <solaris_user> why i cant find this book :( [08:18:51] <jmcp_> silly_girl22: except for the slight problem that my pc doesn't support hotswap pci cards, I would have [08:19:20] <solaris_user> any knoes one site which sell solaris internals 2nd by paypal?!?!?!? [08:19:33] <Error_404> ebay, likely [08:19:55] <solaris_user> 0)0 [08:20:46] <solaris_user> #find /ebay -book "^Solaris Internals*" -print [08:21:02] <Error_404> finally, some not-crap throughput from this drive [08:25:48] <solaris_user> btw im waiting our new storage to install my first zfs pool :) [08:26:28] <Error_404> yeah... don't give it full drives unless you've got a decent SATA chip [08:26:33] <Error_404> and a decent motherboard [08:26:43] * Error_404 found that out the hard way [08:27:41] <solaris_user> Error_404 : well my college buy one 20 SCSI 3 HD pool storage :) I does not has problem with bus, I guess [08:28:50] <solaris_user> bbougth [08:28:59] <Error_404> no, my ghetto SATA setup with cheapass motherboard evidently has problems with freezing on EFI disk labels [08:29:06] <solaris_user> bought [08:29:46] <solaris_user> :( I need Solaris Internals 2nd book -_- anyone has mercy [08:30:00] <Error_404> cheapest i saw it for was amazon [08:30:12] <solaris_user> 8 weeks to delivery -_- [08:30:15] <solaris_user> live in brazil [08:30:51] <Error_404> amazon.br ? [08:31:30] <Error_404> oh, doesn't exist [08:31:33] <solaris_user> :) well i will tell one little history -_- we does not have this "tecnology" here :) [08:31:48] <Error_404> that's wierd, brazil's a big country, why no amazon brazil? [08:32:07] <solaris_user> i really dont know [08:32:12] <Error_404> technology aside, it's not a country of illiterates [08:32:27] <solaris_user> yeap [08:32:40] <Error_404> there's *got* to be enough people who want books for amazon.br to do well enough to justify a server in a rack somewhere [08:33:09] <jmcp_> solaris_user: have you tried phptr.com instead of amazon? I bought my copy from them and it took 1 week to arrive [08:33:36] <solaris_user> nope let me see 1s.............. [08:33:49] <solaris_user> http://www.amazon.com/Solaris-TM-Internals-OpenSolaris-Architecture/dp/0131482092/ref=pd_rhf_p_1/104-5673807-3815164?ie=UTF8 <-- need this! [08:34:23] *** tsoome has joined #opensolaris [08:34:24] <jmcp_> http://www.phptr.com/cart/ordering.asp?rl=1 [08:34:37] <jmcp_> http://www.phptr.com/bookstore/product.asp?isbn=0131482092&rl=1 [08:34:38] <solaris_user> thx [08:35:16] <jmcp_> don't forget vol2 http://www.phptr.com/title/0131568191 [08:35:49] <solaris_user> I have that :) [08:37:04] <solaris_user> bought ebook .pdf 2 months ago [08:37:32] *** Xh4 has quit IRC [08:41:03] <Sparcioni> is there something like netstat -l ? how can I know if a port is free or occupied ? in solaris there is no -l for netstat [08:41:31] <solaris_user> or man :) [08:43:27] <solaris_user> Xorg doesnt like me -_- [08:44:18] <solaris_user> I prefer OW with Xsun [08:45:35] *** pseudo_ has joined #opensolaris [08:45:39] <Error_404> hmm... i wonder what'd happen if i put this SATA controller in RAID mode & left only a single drive on it [08:45:53] *** mazon is now known as Mazon [08:47:33] <Error_404> nope, that didn't work [08:47:34] <solaris_user> question, has some management tool for ipfilter ( like sunscreen/ solaris 9 ) [08:47:37] <solaris_user> ? [08:49:37] <Sparcioni> jmcp_, I am trying : dtrace -n'java:::' -p 100978 [08:49:45] <Sparcioni> and getting: dtrace: invalid probe specifier java:::: probe description java::: does not match any probes [08:50:12] <Sparcioni> should there be someting instead of ::: [08:50:12] <solaris_user> cant help you with java 0)0 [08:50:36] <solaris_user> miss [08:50:55] *** Dar_HOME is now known as Dar [08:50:58] <Sparcioni> truss -p pid shows lots of info [08:51:13] <solaris_user> truss the best debug ever :) [08:51:20] <Sparcioni> is pollsys a sys call? [08:51:20] <solaris_user> and easy to use :) [08:51:28] <Sparcioni> I see many pollysys [08:52:10] <andersmo> solaris_user: if you like truss, you'll love dtrace. =) [08:52:11] <Sparcioni> but can truss show memory statistics? [08:52:33] <andersmo> Sparcioni: what are you trying to probe? [08:52:34] <solaris_user> :) [08:52:34] <Sparcioni> andersmo, any idea what the dtrace error i have is about ? [08:52:44] <solaris_user> but i still learning dtrace [08:52:50] <Sparcioni> andersmo, this is a test app which sends UDP packets [08:52:53] <Sparcioni> in a loop [08:52:59] <andersmo> Sparcioni: well, there is simply no provider called "java". [08:53:23] <andersmo> The built-in provider in JDK6 is called hotspot. [08:53:24] <Sparcioni> andersmo, what does this mean ? what should I do? [08:53:50] <jmcp_> andersmo: so when team dtrace refers to the java provider, what provider prefix are they meaning? [08:53:53] <solaris_user> cant help with java provider -_- i really doesnt know [08:54:33] <Sparcioni> I have java 1.5.0_07 [08:54:50] <Sparcioni> on solaris sparc (NIAGRA) [08:56:39] <andersmo> Sparcioni: There's a provider called dvm for java 1.5 [08:56:48] <jmcp_> ah, that's what I was wondering [08:56:53] <Sparcioni> andersmo, trying now... [08:57:13] <Sparcioni> dtrace -n'dvm:::' -p 100978 [08:57:14] * jmcp_ wanders home [08:57:15] <andersmo> Howeever, it needs to be explicitly loaded with a command to the jvm during startup. [08:57:16] *** jmcp_ has quit IRC [08:57:17] <Sparcioni> the same error [08:57:26] <Sparcioni> dtrace: invalid probe specifier dvm:::: probe description dvm::: does not match any probes [08:57:39] <Sparcioni> am I doing something incorrect issuing this dtrace? [08:57:42] <andersmo> Sparcioni: take a look at http://www.google.no/url?sa=t&ct=res&cd=4&url=http%3A%2F%2Fjava.sun.com%2Fj2se%2Freference%2Fwhitepapers%2Fjava-dtrace-whitepaper.pdf&ei=3-YtRYznCpe-wgG0wvnjAQ&sig=__P52DBZD1oZCDYd72-P1el5BKeLo=&sig2=saVo3HpGRicspTAZgx5VPQ#search=%22java%20dtrace%20provider%20dvm%22 [08:57:47] <andersmo> ouch, sorry. [08:57:50] <andersmo> java.sun.com/j2se/reference/whitepapers/java-dtrace-whitepaper.pdf [08:57:57] <andersmo> There's a more sane URL. =) [08:58:00] <Sparcioni> looking [08:58:21] <Sparcioni> can any one try dtracde some java process on his machine ? [08:59:03] <andersmo> Sparcioni: Look at the section called "Adding DTrace Probes to Pre-Mustang Releases Using VM Agents" on page 7. =) [08:59:19] *** slowhog has quit IRC [08:59:25] <andersmo> Or read the whole paper, it's useful stuff for what you're doing now. =) [08:59:39] <Sparcioni> reading ... [08:59:40] <andersmo> Sparcioni: I've dtraced java a lot. Works like a charm. =) [09:00:01] <Sparcioni> andersmo, with dvm? [09:00:16] <andersmo> No, I use a JDK1.6 prerelease to get the hotspot provider. =) [09:00:33] <andersmo> But dvm and hotspot can do pretty much the same stuff. [09:00:51] <Sparcioni> andersmo, thnks [09:01:16] <andersmo> But there's a significant performance hit to some of the probes, and I think the impact is worse in the dvm provider. [09:05:46] <Error_404> forget this disk crap, i'm going to bed [09:05:52] <solaris_user> :) [09:05:56] <solaris_user> gn [09:06:44] <Sparcioni> andersmo, Pre-Mustang on page 7: what is Mustang ? [09:06:53] <Sparcioni> sorry my ignorance [09:06:56] <andersmo> Sparcioni: mustang is java 6. =) [09:07:03] <andersmo> aka jdk 1.6 [09:07:07] <Sparcioni> ok, thnks [09:07:28] <andersmo> Or whatever sun marketing wants to call it. Anyway, you understand. =) [09:08:42] <Sparcioni> :) [09:08:43] <solaris_user> zones and containers -_- [09:09:03] <solaris_user> sun marketing has some sexual problens with names [09:10:31] *** Jiko has quit IRC [09:11:37] <solaris_user> i guess ill subscribe safari books online [09:17:36] *** uzi__ is now known as uzi [09:25:11] *** Cyrille has joined #opensolaris [09:25:29] *** jengelh has joined #opensolaris [09:26:33] <Sparcioni> andersmo, a question : I installed it and set environment vars according to instructions an still getting dtrace: invalid probe specifier dvm:::: probe description dvm::: does not match any probes [09:26:49] <Sparcioni> should I do something else or are there any default probes ? [09:27:36] <Sparcioni> also with dvmti the same [09:28:03] <Sparcioni> JAVA_TOOL_OPTIONS is -Xrundvmti:all [09:28:30] <Sparcioni> env | grep LD_L [09:28:31] <Sparcioni> LD_LIBRARY_PATH_64=/opt/work/down/dvm/build/sparcv9/lib [09:28:32] <Sparcioni> LD_LIBRARY_PATH_32=/opt/work/down/dvm/build/sparc/lib [09:28:48] <Sparcioni> oH NO! [09:28:52] <Sparcioni> i FORGOT [09:28:57] <Sparcioni> java -Xrundvm[tp]i[:options] .. [09:29:45] *** Kernel86|Laptop has joined #OpenSolaris [09:32:17] <Sparcioni> did not help: also java -Xrundvmti className [09:32:32] <Sparcioni> and then: dtrace -n'dvmti:::' -p 101074 [09:32:39] <Sparcioni> gave dtrace: invalid probe specifier dvmti:::: probe description dvmti::: does not match any probes [09:32:44] <asyd> \_o< [09:32:52] <Sparcioni> ? [09:33:57] *** solaris_user has quit IRC [09:36:11] <andersmo> Sparcioni: I think the dvm provider name during runtime is "dvm12345", where 12345 is the pid of the JVM. [09:36:31] <Sparcioni> trying... [09:37:28] <Sparcioni> wow ! [09:37:33] <Sparcioni> works !!!!!!!!!!!! [09:37:35] <Sparcioni> thnks [09:37:36] <andersmo> =) [09:37:54] <Sparcioni> they should add it in that page https://solaris10-dtrace-vm-agents.dev.java.net/ [09:43:42] *** pseudo_ has quit IRC [09:47:30] *** Elendal has joined #opensolaris [10:07:11] *** Sparcioni has quit IRC [10:10:00] *** peteh has joined #opensolaris [10:10:04] *** Gman has quit IRC [10:10:05] *** alanc_away has quit IRC [10:10:58] *** alanc_away has joined #opensolaris [10:12:05] *** jamesd_ has joined #opensolaris [10:12:05] *** jamesd has quit IRC [10:15:19] *** Yamaraj has joined #opensolaris [10:25:53] *** peteh has quit IRC [10:27:07] *** coffman has joined #opensolaris [10:43:20] <kimc> good morning [10:43:40] *** bengtf has joined #opensolaris [10:46:03] *** elflord has quit IRC [10:55:55] * Tpenta is now ready for when stevel does the b50 drop. I have the nd encumbered bins built and ready to be copied onto dlc once he has put the rest up there :) [11:08:06] *** tsoome has quit IRC [11:10:52] *** MattMan has joined #opensolaris [11:16:59] *** Yamaraj has quit IRC [11:20:36] *** Capricorn^80 has joined #opensolaris [11:20:55] <Capricorn^80> need help in configuring NS 2 [11:21:08] <asyd> need money [11:21:50] <Capricorn^80> GOD will give u :) [11:22:22] * quasi hands asyd EUR .20 - don't spend it all in one place ;) [11:22:32] <asyd> :) [11:22:55] <asyd> quasi: the english version of the howto for the dedicated server should be available in few days [11:23:15] <jengelh> you can hardly spend .20 [11:24:44] <jengelh> Even http://de.wikipedia.org/wiki/Bild:KinderSchokolade.jpg costs already .35 [11:25:16] <asyd> :) [11:28:23] <Cyrille> you can buy a few minutes in some parking meters, maybe. [11:28:51] <sickness> url url [11:29:35] <asyd> http://www.epinal.fr/images/images_viepratique/parcmetre.jpg [11:29:36] <asyd> \o/ [11:30:35] <sickness> url of the howto! :P [11:30:50] <asyd> :)) [11:30:55] <Vanuatoo> I heard solaris 10 update 3 is scheduled for november, is it true? [11:31:03] <asyd> unfortunetaly no [11:31:08] <Cyrille> I'd include the wikipedia link to "parking_meters", but I'm afraid it might explain it was something invented in 1562 by Winston Churchill to measure the dimensions of parking spots... [11:31:09] <asyd> sickness: http://guses.org/home/projets/dedibox [11:31:19] <asyd> Cyrille: any time to translate it ? :P [11:32:19] <sickness> tnx [11:32:20] <Cyrille> asyd: I could do that. [11:32:31] <sickness> :) [11:33:07] <quasi> asyd: excellent [11:33:09] <Cyrille> asyd: was that a serious question? [11:33:39] <asyd> Cyrille: sure. [11:34:07] <Cyrille> asyd: let me see what I can do. Is there a format you'd prefer? [11:34:28] <asyd> I can send you the source (a flat utf8 file) to the page [11:34:59] <Cyrille> I can access the source, actually, from that link you sent. I could keep that format. [11:35:42] <asyd> perfect [11:36:45] <asyd> don't bother you with french mistakes [11:38:53] <quasi> asyd: do you know any details of what sort of machine your c7 2GHz is? [11:39:27] <asyd> what kind of details ? [11:40:09] <asyd> http://kaoru.asyd.net/~asyd/dedibox-solaris.txt kernel in verbose mode + prtdiag -v + prtconf -v [11:40:10] <quasi> I just haven't seen a 2GHz c7 processor before [11:40:53] <asyd> well, I think it is more or less a special processor made by VIA for Dedibox [11:41:24] <quasi> ah [11:43:07] <asyd> dedibox is a subdivision of the second biggest ISP in france, and the more innovative one. They sold more than 6'000 servers in ~6 months [11:44:06] <quasi> that's not bad at all [11:45:48] <asyd> yeah, I have one running linux for 4 months, it's very nice, it's wonderful to have 100mb/s on internet w/o be restricted [11:50:27] *** trs81 has quit IRC [11:51:12] <coffman> in germany u can get a root server from hetzner with a 3700+ amd 64 1gb ram 2x160gb sata hds 100mbit flat - all for 39 euros per month [11:51:36] *** bnitz has joined #opensolaris [11:51:45] *** |ReIkO| has joined #opensolaris [11:51:47] <coffman> i would love to try pulling solaris on that :) [11:51:50] <|ReIkO|> morning all [11:56:07] <|ReIkO|> how can I add a new auth method build by me to SunSSH ? [11:56:18] <twincest> write a PAM module [11:57:54] <|ReIkO|> that I add something like other account required pam_projects.so.1 [11:58:42] <|ReIkO|> but than how can I call this method? [11:58:54] <twincest> ssh will do it for you based on the pam configuration [12:00:12] <asyd> coffman: it's nice too. [12:00:34] <asyd> coffman: well, if you ask some admin to get a serial access port via telnet, I can help you to install a solaris from a linux :) [12:03:05] <quasi> asyd: so your method still needs serial access? [12:03:12] *** tfb has quit IRC [12:03:14] <asyd> quasi: for the first install yes [12:03:21] <asyd> to determine the bootpath [12:03:32] <asyd> but once the image is ready, other install doesn't require a serial access [12:04:04] <asyd> I still need to check if I can pass the bootpath variable as kernel argument instead used a flat file [12:04:28] <twincest> you can, use -B [12:04:30] <quasi> asyd: so if you were patient, you could just try a bunch of different paths? [12:04:35] <asyd> twincest: ok, nice [12:04:40] <asyd> quasi: yup [12:04:43] <quasi> yeah, that is doable in grub [12:04:48] <twincest> the grub documentation explains the exact variable, i forget if it's also called bootpath there [12:04:56] <asyd> but anyway, I made the first one, so I think it will be very more easy for me now [12:06:19] *** ssam has joined #opensolaris [12:06:44] * quasi would be very interested in getting things going on one of those hetzner machines as well [12:07:41] <quasi> asyd: actually, finding the pci ids and from that the path of sata should be doable on linux with lspci -vvvv [12:07:55] <asyd> yeah that exactly what I'm thinking [12:08:20] <asyd> hmm [12:08:35] <asyd> from linux : 0000:00:0f.0 IDE interface: [..[ [12:08:39] *** ssam has quit IRC [12:08:40] <asyd> which give bootpath=/pci@0,0/pci-ide@f/ide@0/cmdk@0,0:a [12:09:34] *** ssam has joined #opensolaris [12:09:39] <asyd> do you know the chipset of the network card on hetzner servers ? [12:10:02] <quasi> I can find out [12:10:49] *** calumb has joined #opensolaris [12:11:41] <asyd> |ReIkO|: btw, if you thinking about x509 authentication with openssh, a patch is available. [12:15:41] <quasi> asyd: http://www.ellegaard.dk/klaus/lspci-vvv.txt [12:15:53] <asyd> realtek ! nice [12:16:06] <twincest> there's two words you don't often hear together [12:16:13] <asyd> 0000:00:0f.0 RAID bus controller: VIA Technologies, [12:16:22] *** kimc has left #opensolaris [12:16:25] <asyd> funny, seem it's the same chipset [12:17:01] <asyd> twincest: :) [12:17:48] <|ReIkO|> where asyd? [12:17:57] <asyd> google://roumen petrov/ [12:18:24] <asyd> quasi: well, do you have 254 cylinders free ? ;p [12:18:44] <|ReIkO|> thanks asyd [12:19:55] *** ssam has quit IRC [12:31:20] <asyd> hmm am I supposed to see something related to IPMI on dmidecode output ? [12:39:36] *** smferris has quit IRC [12:51:03] *** Capricorn^80 has quit IRC [13:28:34] *** yarihm has joined #OpenSolaris [13:33:24] <timeless> silly question, solaris normally uses pkgadd to install software, right? [13:33:36] <Cyrille> silly answer, yes [13:33:50] <timeless> ok, user error [13:33:58] * timeless botched the syntax for pkgadd, missed the -d [13:34:11] <Cyrille> common mistake. [13:38:11] * timeless tries to figure out where to go for help w/ blastwave pkg-get [13:38:12] *** MattMan has quit IRC [13:38:17] *** MattMan_ has joined #opensolaris [13:39:32] <timeless> ok, so i have this /opt/csw and /var/pkg-get, would it be unreasonable for me to make a zfs volume and symlink those two directories into it so that i can version those indepndently of everything else? [13:39:53] *** |ReIkO| has quit IRC [13:41:44] *** Jiko has joined #opensolaris [13:42:38] <quasi> asyd: I don't have one of the hetzner boxes yet - the lspci output was from a friends box [13:44:23] <asyd> ok [13:45:58] *** Tpent1 has joined #opensolaris [13:49:12] * timeless sighs [13:49:33] <timeless> the solaris archive manager doesn't seem to support .deb even though it's configured to be the default handler for them [13:52:17] *** aliquis has joined #opensolaris [13:53:29] *** solarisjon has joined #opensolaris [13:54:48] <aliquis> When the installer installs grub is it installed on MBR or in the begining of the Solaris partition? Somehow I've lost mine and I would like to be able to get it back with the same settings as easy as possible [14:02:11] *** calumb is now known as calLNCH [14:03:43] <PerterB> I _think_ by default in the Solaris partition as there's a seperate -m flag to installgrub to patch the MBR (so I'm assuming the default is to leave the MBR as-is to boot whichever partition is marked active) [14:06:43] *** dunc_ has quit IRC [14:07:33] *** sparc-kly has joined #opensolaris [14:07:34] *** ChanServ sets mode: +o sparc-kly [14:08:29] *** tsoome has joined #opensolaris [14:08:56] <asyd> ok, guys, thanks to Cyrille [14:08:58] <asyd> http://guses.org/home/projets/dedibox.en [14:10:27] <jengelh> asyd : No cooling racks? [14:10:51] <asyd> ?! [14:12:41] <jengelh> like these 'closed racks' http://www.sgi.com/features/2004/oct/columbia/images/lg_receding_vw.jpg [14:13:26] <asyd> I don't care about the datacenter, it's not image. Picture are here just for fun :) [14:13:45] <jengelh> well looks like dediboxes are just on rails rather than boxes [14:14:35] <asyd> maybe. The more important thing for me that dediboxes are very cheap and it's now possible to use solaris ;) [14:16:30] *** laca has joined #opensolaris [14:17:08] * timeless just wants some way to get apt to half work on opensolaris :( [14:17:22] <PerterB> less cheap if you don't have a french bank account... $59.95 from that bestofdata place vs EUR29.99 direct from dedibox [14:17:26] <Cyrille> talking about cheapness, I don't know if you've noticed that bestofdata applies a rather generous dollar to euro conversion rate. [14:17:31] <asyd> PerterB: yeah I know :/ [14:17:32] <Cyrille> my point [14:17:55] <timeless> heh [14:18:21] <asyd> though, I'm talking with to dedibox to have a free box for our user group [14:18:53] <PerterB> then again, the way the dollar has been going, EUR30 might end up being $60 ;) [14:18:57] <asyd> and anyway if some here need some hosting (especially related to (open)solaris) feel free to ask [14:20:51] *** oooo has joined #opensolaris [14:27:25] *** lloy0076 has joined #opensolaris [14:27:36] * lloy0076 ponder [14:28:58] <lloy0076> I'm trying to get OpenSolaris onto my main machine. The only way I can think of it is to: 1. Install OpenSolaris on the first 60G; 2. Install NexentaOS on a partition after that as small as I can get it 3. Transfer the Sun compiler and the via-rhine drivers to the open solaris install 4. Build the ethernet drivers for Open Solaris 5. Use OpenSolaris [14:29:33] <lloy0076> This is primarily because I'll fall over my original: I need the network to get the drivers and compiler; but I won't have the network if I don't have the drivers. [14:29:35] * lloy0076 sigh [14:29:57] <lloy0076> [and no I don't have a CD burner] [14:30:40] <asyd> you probably have USB [14:30:53] <asyd> and an usb key with fat32 fs [14:31:13] <lloy0076> Not at the moment, but I could get one. They're cheap enough. [14:31:14] <PerterB> how about install opensolaris, then use a livecd to ftp on the bits you need [14:31:24] <PerterB> oh, no cd burner duh (sorry) [14:31:31] <lloy0076> Ah, PeterB is correct. [14:31:40] <lloy0076> I *do* have the nexenta live CD [14:31:43] * lloy0076 thwaps self [14:32:05] <lloy0076> I remember the reason that didn't work before was because I couldn't work out Solaris' partition/slice/format naming. [14:32:06] <jteo> saved again! [14:32:39] *** calLNCH has quit IRC [14:32:42] <tsoome> pc's sucks.... [14:32:48] <timeless> hrm, if i had a nexenta livecd, would it be possible for me to pull enough from its apt in order to run it? [14:34:36] <lloy0076> Incidentally, I suppose OpenSolaris is able to burn DVDs and such... [14:34:41] <timeless> yes [14:34:44] <timeless> it works too [14:35:05] *** phearless has quit IRC [14:35:08] <timeless> iirc that's how i made my snv_43 dvd [14:43:18] *** jsubl2 has joined #opensolaris [14:45:00] <lloy0076> Anyway, off I go into the sunset (so to speak)... [14:45:04] *** lloy0076 has quit IRC [14:46:27] *** ShadowHntr has joined #opensolaris [15:01:56] *** laca is now known as lacaAFK [15:10:23] *** vmhobbes has joined #OpenSolaris [15:10:42] *** dunc_ has joined #opensolaris [15:14:22] *** nachox has joined #opensolaris [15:15:46] *** glagasse has joined #opensolaris [15:17:20] <aliquis> PerterB: ok, so maybe some program just changed active partition to the win xp one [15:17:32] <aliquis> PerterB: so in that case it's probably enough to just switch active one to solaris, thanks [15:18:21] <aliquis> PerterB: I was trying to install mac os x on my pc aswell and then i tried partition magics boot magic and after that i only got the win xp menu, but i suppose boot magic used win xps menu/boot loader to handle boot from other partitions then [15:19:57] *** triplah has quit IRC [15:21:47] <Dr_Jekyl1> c [15:22:13] *** Dr_Jekyl1 has left #opensolaris [15:25:17] <PerterB> aliquis: yeah, it can be pretty confusing with multiple OSes depending on whether some of them patch the MBR [15:31:12] *** ShadowHntr has quit IRC [15:33:36] *** mkhl has joined #opensolaris [15:34:44] <aliquis> PerterB: yeah, i guess i could just have tried to change active partition first, but in case windows wouldn't boot then either .. ;), also i actually run mac os x on my regular pc atm so i was to lazy to reboot ;/ [15:37:09] <aliquis> http://img140.imageshack.us/my.php?image=picture1gl9.png ;) [15:39:41] *** soultan has joined #opensolaris [15:40:28] *** lacaAFK is now known as laca [15:42:39] *** jteo has quit IRC [15:47:37] *** jlc has joined #opensolaris [15:49:41] *** mrdeviant has joined #opensolaris [15:53:45] <soultan> hello i'm an art student doing some research on your wonderful operating system [15:53:50] <soultan> would anybody mind showing me their system startup message? as well as other text output that might show me how it's like to use opensolaris? [15:55:14] <asyd> that's sux a lot it's not yet possible to have multiple routing tables in sol10 @#$2 [15:56:42] <hile_> eh? [15:57:33] <asyd> yeah ? ;p [15:58:09] <PerterB> for what purpose? [15:58:32] <hile_> my thoughts exactly [15:59:01] <asyd> ah, well [15:59:04] <soultan> i'm curious [15:59:07] <soultan> dmesg > file.txt [15:59:13] <asyd> hile_: PerterB you talk to me I guess ? [15:59:14] <soultan> or just paste it to paste.husk.org [15:59:15] <soultan> pelase guys [15:59:22] <PerterB> asyd: yup [15:59:39] <mrdeviant> soultan, solaris is actually pretty quiet on boot because of changes made in Solaris 10 (SMF) [15:59:40] <asyd> so, guess you have a Sun directly on the internet [15:59:54] <soultan> i see i see [15:59:57] *** smferris has joined #opensolaris [15:59:57] <asyd> you want the global zone reachable only from LAN [16:00:05] <asyd> w/o public IP [16:00:12] <asyd> so you need one default route to the LAN router [16:00:20] <soultan> i am really just interested in gathering as much text output as possible so i can form an idea of how information is structured ... [16:00:36] <asyd> but you want some local zones reachable from both lan Internet / LAN [16:00:56] <asyd> with two physical networks [16:01:09] <quasi> asyd: I do that all the time [16:01:10] <ofu> soultan: /var/adm/messages, /var/svc/log/* [16:01:11] <asyd> actually, it's possible but with a ugly hack (define two default router, one in LAN subnet) [16:01:31] <asyd> soultan: http://kaoru.asyd.net/~asyd/dedibox-solaris.txt [16:01:36] <quasi> asyd: no 2 default routers [16:02:07] <asyd> hmm [16:02:16] <asyd> few seconds, let me find my very old opensolaris post [16:02:35] <asyd> http://asyd.net/home/docs/solaris/net_zones [16:02:40] <asyd> you need to make something like that, right ? [16:03:43] <soultan> wow [16:03:44] <soultan> asyd [16:03:45] <soultan> thanks a lot man [16:12:34] <hile_> asyd -- i am probably missing something, but I still don't see why you would need two default routes. If you need to do something cute, I'd think that you'd configure your routing protocols upstream from the individual hosts to accomplish that [16:13:07] <asyd> hile_: hmm [16:13:16] * quasi agrees with hile_ - one default route is enough [16:13:17] *** sommerfeld has joined #opensolaris [16:13:17] *** ChanServ sets mode: +o sommerfeld [16:13:20] <asyd> ?! [16:13:25] <asyd> well, guess the following case : [16:13:51] <asyd> global zone with 192.168.1.240/24, you want (it's a requirement) the default route for global zone is 192.168.1.1 [16:14:53] <asyd> you have a local zone with a public address on another nic, like 9.2.2.2, and you want that zone use 9.2.2.1 as default router [16:15:47] <AbeFroman> sunw rallying... it's up .04! [16:16:13] <quasi> asyd: no [16:17:05] <quasi> asyd: in the global zone you define 9.2.2.x and 192.168.1.24x because you need something configured on each net in the global zone [16:17:32] <quasi> asyd: for local zones you give them an ip on one or the other networks and they'll use that [16:17:39] <asyd> yeah, but I found that ugly. [16:18:13] <quasi> asyd: it is how tcp/ip works [16:18:17] <soultan> asyd: are there other outputs that may be important as well? top for instance or something? [16:18:19] <asyd> no [16:18:24] <quasi> asyd: nothing ugly about it [16:18:55] *** Mazon is now known as mazon [16:19:09] <asyd> quasi: but you need to setup a public IP on the global zone [16:19:38] <asyd> anyway [16:19:49] <asyd> I know multiple routing table is planned [16:19:52] <quasi> asyd: yes, that you do - it is part of the requirements [16:20:26] *** loke has joined #opensolaris [16:20:34] <soultan> hehe [16:20:34] <quasi> asyd: "the global zone needs an ip on each network..." - it is in the docs [16:20:36] <soultan> oh wait [16:20:39] <soultan> asyd, please, one last thing [16:20:44] <soultan> some output showing zfs in action! [16:20:47] <soultan> please man [16:20:58] <asyd> soultan: http://kaoru.asyd.net/~asyd/zfs/zfs.html [16:21:01] * aliquis smells school assignment [16:21:07] <asyd> quasi: there is a hack [16:21:14] <asyd> set an IP on the global zone [16:21:17] <asyd> add the route [16:21:18] <asyd> remove the IP [16:21:21] <asyd> boot the zone [16:21:27] <aliquis> soultan: just install it yourself and look? [16:21:38] <soultan> i don't have a spare machine to do it on [16:21:43] <soultan> i'm working fulltime now [16:21:44] <soultan> maybe in a bit [16:21:51] <quasi> asyd: and pick up the pieces every time you have to reboot the global zone [16:21:56] <quasi> asyd: no thanks ;) [16:22:09] <soultan> thanks asyd [16:22:12] <asyd> that's why I _Want_ multiple routing table. [16:22:25] <asyd> and for isolation [16:22:48] <twincest> i heard solaris is getting policy routing [16:22:57] <twincest> which will do what you want [16:23:05] <quasi> asyd: sure, grab trusted extensions... ;) [16:23:25] <asyd> yeah, u3 will be very nice [16:23:52] *** trs81 has joined #opensolaris [16:24:08] * hile_ looking more toward U4 [16:24:18] <hile_> the one BF that i'm waiting on is scheduled for u4 [16:24:37] <asyd> which feature you're waiting ? [16:24:44] * quasi was in the middle of installing u3 when he got distracted [16:24:47] <asyd> :)) [16:24:59] <asyd> by a stupid guy who tried to make strange thing with networking ? ;p [16:25:29] <hile_> krb5 header [16:25:30] <hile_> s [16:25:38] <asyd> ahhh I finally found my old post [16:25:39] <asyd> http://forum.sun.com/jive/thread.jspa?forumID=299&threadID=75619 [16:26:29] <asyd> don't be rude, I was young and stupid [16:26:32] <asyd> now I'm just stupid [16:27:05] *** tsoome has quit IRC [16:27:37] <twincest> hile: gssapi! [16:27:48] <hile_> twincest: ordinarily yes [16:27:58] <hile_> not possible for what I'm working with though [16:28:14] <hile_> OpenAFS in this case [16:28:16] <soultan> lol [16:28:27] <asyd> :) [16:28:31] <asyd> nice challenge [16:29:29] <hile_> because it (at this time at least -- i'll have to ask libkeiser if usin gssapi overhaul is at all planned) uses krb5 functions internally, it needs the headers [16:30:07] <mrdeviant> rxgk is planned, and the spec is (mostly) written, but last i heard no one was actively writing the implementation. [16:31:22] * hile_ likely lacks the requisite clue to do so. [16:31:35] <asyd> rxgk ? [16:31:42] <hile_> you guys still looking at NFSv4 to replace DCE/DFS? [16:31:51] <mrdeviant> an Rx security plugin based on gssapi [16:32:00] <asyd> ok [16:32:00] <mrdeviant> hile_, most likely [16:32:24] <mrdeviant> it's actually looking like nfsv4 backed with gpfs will be the solution [16:32:38] <hile_> why does that scare me a bit? [16:32:38] <mrdeviant> but we've had some gpfs hiccups lately [16:32:53] <hile_> gpfs sounds pretty cool, but I haven't got my hands on any filesets to check it out [16:33:18] <janus_man> /msg nickserv set unfiltered on [16:39:32] *** Dar is now known as Dar_HOME [16:39:42] *** NilsAN has quit IRC [16:45:52] * elektronkind wonders when openafs 1.4.2 will finally be cut [16:46:56] <mrdeviant> one of the gatekeepers' problems is lack of feedback from testers [16:47:02] *** elflord has joined #opensolaris [16:47:10] <mrdeviant> if you haven't found any bugs, drop them an email and let them know that. [16:58:20] *** Kush- has joined #opensolaris [16:59:06] *** stevel has joined #opensolaris [16:59:06] *** ChanServ sets mode: +o stevel [17:02:59] *** dunc_ has quit IRC [17:05:14] *** dunc_ has joined #opensolaris [17:06:22] *** Berny has quit IRC [17:07:42] *** schily___ has joined #opensolaris [17:16:14] *** schily__ has quit IRC [17:17:56] *** jamesb has joined #opensolaris [17:19:41] *** mazon is now known as Mazon [17:21:31] *** LordKing has joined #opensolaris [17:29:02] *** vmhobbes has quit IRC [17:30:17] *** hell` has joined #opensolaris [17:31:26] *** yarihm has quit IRC [17:31:29] <jamesb> hi guys, anyone know much about RAID-Z? [17:31:41] <sommerfeld> just a little [17:31:53] <jamesb> :) [17:32:12] <sommerfeld> ask your question.. [17:33:03] <jamesb> From what I've read, because of dynamic stripes, metadata tells you which disks contain portions of the stripe... [17:33:26] <jamesb> so, it presumably wouldn't be too hard to(?) to implement RAID-Z with disks which aren't the same size [17:33:43] <jamesb> or are there technical reasons why this is difficult? [17:34:21] *** jteo has joined #opensolaris [17:34:31] <jteo> re. [17:35:53] *** axisys has joined #opensolaris [17:35:58] <stevel> morning jteo [17:36:10] <jteo> stevel: morning. :) [17:36:53] <sommerfeld> so, the current state of raid-z is that you either have full-width stripes or mirroring of logically adjacent blocks for sub-stripe writes. [17:37:25] <jamesb> ok [17:37:33] <sommerfeld> there are some reserved bits in the block pointer (called "grid bits"), the use of which hasn't been completely defined yet. [17:38:04] *** calumb has joined #opensolaris [17:38:05] <sommerfeld> so conceivably those could be used to handle "the tail end of disks 3, 4 and 5 is missing" case [17:38:52] *** vmhobbes has joined #OpenSolaris [17:42:46] <axisys> is there a quick way to find the absoulte device path of a network interface? [17:42:58] *** jlc has quit IRC [17:44:04] <axisys> got it [17:45:24] <ofu> grep $interface /etc/path_to_inst [17:47:09] <axisys> ofu: cool [17:47:34] *** bank has joined #opensolaris [17:47:52] <bank> hi [17:49:49] <jamesb> thanks for the info sommerfeld [17:50:39] *** oooo has quit IRC [17:51:28] <bank> :D hi jamesb [17:51:29] *** stevel changes topic to "opensolaris.org down for 15 minutes | Latest releases: SXCR: 49, ON build: 50, ON nightly: 20061009" [17:51:54] <jamesb> hi [18:01:58] *** mwaterl has joined #opensolaris [18:05:01] <gdamore> hi * [18:06:14] *** slowhog has joined #opensolaris [18:08:45] <bank> hi gdamore [18:09:25] <gdamore> installing b49 now. so is SunVTS 6.3 actually _on_ the SXCR DVD media? [18:09:40] *** jlc has joined #opensolaris [18:10:29] *** jfndi has joined #opensolaris [18:11:49] <bank> it would be great if there are opensolaris podcast to listen. [18:12:03] <gdamore> i'm very, very surprised to find that suninstall still has motif widget applications (the disk partitioner is a motif app) and it uses mwm or dtwm instead of something like metacity. [18:14:07] <sstallion> well [18:14:16] <sstallion> solaris has been around for quite some time :) [18:15:05] <gdamore> yes, but it would actually make the miniroot smaller, i think, if they could just use _one_ widget set during install. :-) [18:15:25] <sstallion> sure, but it takes a good bit of time to update bits of the installer [18:15:27] <gdamore> i still remember when this was all done in olwm. :-) [18:15:41] <sstallion> yup [18:15:50] * elektronkind fenagles a way to get drive temperatures from solaris [18:16:11] <gdamore> actually, i remember when installation was done in a text window from _tape_ in SunOS 4.x days. :-) [18:17:15] <mrdeviant> elektronkind, smartctl ? [18:17:16] *** stevel changes topic to "Latest releases: SXCR: 49, ON build: 50, ON nightly: 20061009" [18:17:39] <elektronkind> mrdeviant: via libdiskmgt stats provider [18:18:05] <elektronkind> mrdeviant: unfortunately libdiskmgt is not a public interface, so its header file is missing from /usr/include [18:18:18] <elektronkind> and it sure as hell aint documented. [18:19:12] <elektronkind> mrdeviant: http://pastebin.ca/199541 [18:19:17] <sstallion> hrmm. when will sun finishing moving sun studio over to netbeans? [18:19:39] <sstallion> well, netbeans 5.x :P [18:20:01] <mrdeviant> right, or you could just use smartctl and have less pain [18:20:43] <elektronkind> smartctl isn't on solaris 10, and I'm making a general disk/buss info tool. [18:21:05] <mrdeviant> it's trivial to add to s10 [18:21:11] <bank> sun studio will integrated into netbeans? [18:21:44] <bank> netbeans C/C++ is not good enough. [18:21:48] <sstallion> bank: eventually. take a look at the c/c++ pack [18:21:50] <gdamore> i still can't bring myself to use a full IDE. I use emacs, but I don't use any of the builtin compilation or debug features. [18:22:13] <sstallion> its actually somewhat decent, i still like studio better, but its starting to show its age [18:22:18] <sstallion> im a vim guy personally [18:22:19] <bank> This week I need to convert C++ data mining algorithm to java. What I do is download Visual studio Express. [18:22:36] <bank> But I try Netbeans C/C++ also. It doesn't have many feature. [18:22:46] <bank> to trace the variables or method call. [18:23:38] <gdamore> back when I was working for Sun, people tried to get to even just use cscope, but I never adapted. I use grep a lot. :-) [18:24:35] <bank> sstallion: I am opening milestone. [18:24:37] <gdamore> (i suspect, that as a result, i have a much better feel for where stuff in source trees is. i can find pretty much anything I need in the Solaris kernel, Sun Ray firmware, or NetBSD kernel trees. :-) [18:25:16] <bank> it seem that it doesn't even version 1.0 yet. [18:25:38] <sstallion> bank: its still beta [18:25:54] *** Kernel86|Laptop has quit IRC [18:25:58] *** Kernel86_ has joined #OpenSolaris [18:26:06] *** nwf_ has joined #opensolaris [18:27:04] <bank> i decide to use netbeans platform to develope rcp base application. by the way. my professor have an ibm academic initialtive and finally force me to use eclipse rcp. [18:27:33] <sstallion> blech [18:27:47] <sstallion> its really tough to beat idea [18:27:54] *** loke has quit IRC [18:27:58] <sstallion> if only theyd create a c/c++ ide [18:30:21] *** Tzoa has joined #opensolaris [18:30:30] *** nwf has quit IRC [18:32:09] *** Tzoa has quit IRC [18:32:14] *** Tzoa has joined #opensolaris [18:38:53] *** MattMan_ has quit IRC [18:39:06] *** nwf_ is now known as nwf [18:39:35] *** sparc-kly has quit IRC [18:45:47] *** jsubl2 has quit IRC [18:47:32] *** jfndi has left #opensolaris [18:48:09] *** sparc-kly has joined #opensolaris [18:48:09] *** ChanServ sets mode: +o sparc-kly [18:54:19] *** mkhl has quit IRC [18:55:36] *** panicou has joined #opensolaris [18:56:31] *** calumb has quit IRC [18:59:50] *** Sieghard has quit IRC [19:00:25] *** LordKing has quit IRC [19:00:47] <panicou> hello, i have a blade 1000 and i was wondering if i could use a thrid-party pci graphix card to get decent 2d perfomance. [19:01:20] *** sparc-kly has quit IRC [19:01:50] *** sparc-kly has joined #opensolaris [19:01:51] *** ChanServ sets mode: +o sparc-kly [19:03:30] *** Cyrille has quit IRC [19:10:36] <gdamore> its hard to get 3rd party graphics cards that work properly on sparc. [19:11:03] <gdamore> there was a company called TECH-SOURCE that used to supply PGX64 boards that would work fine for 2D. [19:11:46] <gdamore> but if all you are doing is 2D, maybe see if you can pick up an XVR-100. they're not too expensive, I think. you have to get a sun-branded one to get the necessary Fcode on the board [19:11:55] *** bnitz has left #opensolaris [19:13:04] <panicou> thx [19:13:14] <gdamore> unfortunately, the sun pfb driver won't initialize a radeon on its own -- it uses OBP to do that. (I have code would coudl be integrated into the pfb driver to fix that, but i doubt sun would want it, as they still make money selling their own radeon boards) [19:13:39] <Auralis> except that the xvr-100 is rather old crap [19:13:54] <coffman> why not get a mac card? [19:13:59] <gdamore> its a radeon 7000. it works fine. it isn't high end 3D, but its quite nice for 2D. [19:14:35] <gdamore> a mac card might well work. but its unknown territory. OBP is _supposed_ to be portable, but experience is that there are often gotchas [19:14:40] <Auralis> panicou: well, what card do you have currently in the machine? [19:15:29] <gdamore> funny thing is, the xvr-100 could actually do pretty nice 3D, if someone would ever write code for it. :-) it wouldn't be top-end, but it would certainly be competitive with some of the older Sun creator 3D hardware [19:16:07] <oxygene> esp mac-cards are troublesome as they tend to include only a binary driver for mac os.. but at least initialization tends to be fcode [19:16:29] <panicou> a creator 3d [19:16:32] <gdamore> we (tadpole) are actually investigating this particular option, because we have XVR-100s in some of our laptops. (or rather, radeon 7000s) [19:16:51] <gdamore> do you just want a card for an extra display, or for DVI? [19:17:05] <gdamore> 'cause for analog work, a creator 3d is still pretty nice. [19:17:06] <panicou> i also have 2 elite3d spare somewhere [19:17:18] <panicou> its okay [19:17:29] <panicou> but it gets tirring for everyday work [19:17:46] <gdamore> what's wrong with it? apart from analog only output? [19:17:47] <Auralis> creator3d is a nice 2d card, the image quality is just so much bettern then that of cheap pc gfx cards [19:17:57] <panicou> well [19:18:12] <panicou> i have problems when i open mozzila with a flash site for example [19:18:19] <panicou> it gets jerky [19:18:31] *** bank has quit IRC [19:18:38] <gdamore> i doubt, rather strongly, that this is a problem with the creator graphics card [19:18:48] <panicou> hmmz [19:18:56] <Auralis> us the gtk1 build of mozilla/firefox, its significantly faster [19:19:08] <panicou> well will try [19:19:17] <panicou> before getting an other card for sure [19:19:34] <Auralis> the problem is not the card, its gtk2 [19:19:53] <panicou> i would very much like to move to the blade as my everyday machine, but these kind of things hild me back to my windows pc [19:19:58] <gdamore> the thing is, for 2D stuff, all these boards just use a memory mapped framebuffer. nothing really terribly exciting. the only differences will be in bus speed, for most case. [19:20:51] <gdamore> that said, the xvr-100 has nice digital output support. and you can use it with xinerama, which is always a plus. [19:24:34] <gdamore> btw, if you can get a radeon 7000 for MAC, that has fcode on it, it _might_ well work with pfb. but its critical that it be a 7k. other radeon boards will -not- work. [19:25:21] <gdamore> you will lose 3D OpenGL acceleration that you get with creator 3d, though [19:25:40] <Auralis> not that the creator opengl is that hot anyway [19:26:56] <gdamore> no, but its _way_ better than software. [19:29:51] *** panicou has quit IRC [19:32:56] *** nwf has quit IRC [19:34:25] *** solaris-user-1 has joined #opensolaris [19:35:24] *** nachox_ has joined #opensolaris [19:37:11] *** solaris-user-1 has quit IRC [19:41:29] *** nachox has quit IRC [19:42:33] *** dunc_ has quit IRC [19:46:49] *** aliquis has quit IRC [19:51:53] *** adp has left #opensolaris [20:24:58] *** Tzoa has left #opensolaris [20:33:59] *** jsubl2 has joined #opensolaris [20:37:26] *** bondolo has quit IRC [20:39:17] *** nwf has joined #opensolaris [20:39:37] <Error_404> re-learning a keyboard layout is hard [20:42:16] *** pseudo_ has joined #opensolaris [20:43:01] <hile_> switching from a real unix layout back to a crap peeecee keyboard? [20:43:21] <Error_404> no, going back to dvorak [20:43:26] <delewis> hile_: it's not *that* bad. [20:43:35] <Error_404> "real unix layout"? [20:43:54] <hile_> yeah, as in a type 5/6/7 UNIX layout [20:43:56] <delewis> Error_404: a matter of the pinky. [20:43:59] *** dwc_ has joined #opensolaris [20:44:05] <hile_> one with the control key in the right place [20:44:10] <dwc_> !seen stevel [20:44:11] <Drone> stevel is currently online in #opensolaris and last spoke on Thu 12 Oct 2006 15:36 GMT, saying 'morning jteo'. [20:45:37] <Error_404> right place? [20:46:06] <hile_> yes, where control is supposed to be.... above the left shift [20:46:07] <Error_404> oh, in place of the caps lock [20:46:08] <delewis> Error_404: the caps-lock and control keys are reversed. [20:46:26] <delewis> trivial, really. [20:47:04] <Error_404> yeah, that's a pretty trivial change [20:47:08] <hile_> as well has having a compose key for when I need keys not on a US keyboard [20:47:16] <libkeiser> elektronkind: to answer you question from several hours ago, openafs 1.4.2 was cut several days ago. you can grab the tag openafs-stable-1_4_2 out of cvs if you want. we're going through the binary package build process right now. [20:47:32] <delewis> Compose only works when you're in a non-ASCII character set, though :-) [20:47:34] <Error_404> though on the plus side i may have called re-learning a key layout hard a bit prematurely [20:47:35] <bengtf> strangeness when installing: b49 from cds to a u250, when selecting some languages it asks for extra lang cd, but it has already installed languages .... [20:47:39] <delewis> I thought you were a "C" guy, hile :-) [20:48:05] <Error_404> I'm typing more or less normally after ~ a half hour of struggling with it [20:48:08] <hile_> I use 8859-1 [20:48:18] <hile_> en_US.ISO8859-1 [20:48:59] <delewis> ah [20:49:31] *** Sieghard has joined #opensolaris [20:49:38] <Error_404> okay, well that was fun. back to qwerty [20:49:49] <dwc_> dvorak/sun keyboard? [20:50:04] <Error_404> no, it's not a sun keyboard [20:50:12] <Error_404> it's an IBM model M actually [20:50:21] <Error_404> i just pried the keycaps off [20:50:30] <dwc_> nice [20:50:49] <dwc_> "what's your username?" *clickity clackity* [20:51:02] *** metashift has joined #opensolaris [20:51:14] * delewis is typing on a Lexmark model M :-) [20:51:23] <delewis> which is of course, identical to the IBM keyboards (given Lexmark manufactured all of them) [20:51:29] <clee> model M = win [20:51:48] <Error_404> lexmark only manufactured them after 91 [20:52:05] <delewis> were Model Ms around before 91? [20:52:10] <clee> all the best ones were manufactured before 91 [20:52:17] <Error_404> delewis: they certainly were [20:52:18] <clee> delewis: yeah. the two I have at home were made in 86 [20:52:26] <Error_404> the ones that weigh a ton & are made of pure steel [20:52:35] <clee> six pounds, technically. :) [20:52:39] <Error_404> (and have removable cables) [20:52:42] <clee> the newer unicomp ones are only 3lbs [20:52:47] <clee> and they use much cheaper springs :( [20:52:55] <delewis> clee: I doubt there's much of a distinction, though, I do remember seeing variations of the "clicky", especially ones with a few rows of function keys that were meant for IBM AS/400 terminals [20:53:15] <clee> delewis: the biggest difference is just the removable cable, as far as I know. [20:53:26] <clee> delewis: IIRC, the Lexmark models didn't have removable cables. [20:53:29] <Error_404> clee: and the metal underplate [20:53:31] <delewis> clee: mine does [20:53:37] <delewis> I have the AT cable sitting in the drawer beside me :-) [20:53:41] <clee> ah, nice. [20:53:44] <clee> well, in that case, rock on. :) [20:53:51] <Error_404> clee: and the lexmark models have a plastic guard above the circuit board, ostensibly to catch liquids [20:54:38] <clee> weird. [20:54:50] <clee> I've only ever had IBM model Ms [20:54:57] <clee> and the unicomp imitations [20:56:00] *** slowhog has quit IRC [20:58:29] *** _william_ has joined #opensolaris [20:58:59] <_william_> hi [21:00:24] <Error_404> I had a lexmark [21:00:43] <Error_404> traded it & a bag of pot for an IBM to a fool [21:03:47] *** jamesb has quit IRC [21:06:46] <elektronkind> the guy with the ibm got the better deal [21:08:47] *** svoboda has joined #opensolaris [21:10:26] *** jlc has quit IRC [21:14:46] *** Sieghard has quit IRC [21:15:21] *** Sieghard has joined #opensolaris [21:19:00] *** nwf has quit IRC [21:19:14] *** Error_404_ has joined #opensolaris [21:19:56] *** Error_404_ has left #opensolaris [21:20:32] *** Error_404 has quit IRC [21:20:55] *** koolniczka has joined #opensolaris [21:21:01] *** Error_404 has joined #opensolaris [21:21:17] <_william_> hi Error_404 [21:21:31] <_william_> hi koolniczka [21:21:56] <Error_404> hooray, server works again [21:31:05] *** laca is now known as lacaAFK [21:34:10] *** pseudo_ is now known as Xh4 [21:37:57] *** Error_404 has quit IRC [21:37:57] *** yarihm has joined #OpenSolaris [21:45:42] *** metashift has left #opensolaris [21:48:02] *** yarihm has quit IRC [21:51:46] *** tsoome has joined #opensolaris [21:51:52] <hell`> anyone know why a v890 would have 3 NIC's with all the same MAC addresses? [21:52:37] <AbeFroman> eeprom local-mac-address? [21:54:33] <kalif> sparcs usually have the same mac-address on the NICs [21:54:54] <sickness> anyone tried to install dvd::rip on solaris? =) [21:55:30] <hell`> they do? [21:58:15] <kalif> yes [21:58:52] <klocze1> anyone know is it possible setup Solaris NIS client using NIS server on Linux which uses shadow map ? [21:59:07] <hell`> got it thanks [21:59:25] <kalif> klocze1: I did it some years back [21:59:36] *** richlowe has joined #opensolaris [22:00:03] <kalif> lilnux and solrais don't use the same encryption on the passwords [22:00:16] <klocze1> kalif: I have on Linux NIS srv exported shadow map [22:00:43] <elektronkind> kalif: they can. solaris and linux box do 3DES and MD5 [22:00:52] <klocze1> yes but Sol 10 can recognize $1$ MD5 prefix in hash [22:00:54] <elektronkind> s/box/boh [22:01:09] <elektronkind> so can solaris 9 [22:01:18] <klocze1> ok, I use on Linux md5 [22:01:30] <kalif> elektronkind: I couldn't get it to work with soalris 2.6 [22:01:57] <elektronkind> MD5 password hashes weren't supported in solaris until solaris 9 [22:02:04] <kalif> ok [22:02:11] <elektronkind> solaris 2.6 came out in... gosh.. 1996? [22:02:14] <kalif> yes [22:02:19] <kalif> ling time ago :) [22:02:22] <klocze1> but passwd map do not contain crypted passwords (x in each second fileld) [22:02:35] <sickness> yeah, and there was a horrible x86 port too [22:02:43] * kalif runs [22:02:55] <sickness> really really unusable, had almost no drivers for common hardware... [22:02:59] <kalif> don't mention the war :) [22:03:11] <klocze1> kalif: so what next ? :) [22:03:17] <kalif> eh... [22:03:18] <elektronkind> klocze1: because NIS passwords are stored in another database, not in the passwd database. [22:03:50] <kalif> I haven't worked with NIS for years.... [22:04:14] <klocze1> elektronkind: so it is possible only in case when cryptd passwords are axported via passwd.byname map ? [22:05:06] <klocze1> or is it some way for import shadow map on Sol (10) client ? [22:06:33] <elektronkind> what do you get if you "ypcat shadow" on the client? [22:07:06] <klocze1> bash-3.00# ypcat shadow [22:07:06] <klocze1> internal yp server or client error [22:07:21] <elektronkind> well there ya go [22:07:34] <klocze1> "ypcat -x" don't shows shadow map :> [22:07:54] <kalif> but..... [22:07:59] <elektronkind> your shadow.byname map isn't exported [22:08:19] <kalif> does it make sense with shadow passwords and NIS? [22:08:51] <elektronkind> are yu using merged passwd maps on the yp server? [22:09:06] <jamesd_> http://www.sun.com/training/catalog/courses/SA-229-S10.xml?feed=RSS thinks damm... $1400 for training how to be a ZFS administrator.... that is bloody insane... [22:09:23] <kalif> training is big business [22:09:26] <elektronkind> "zfs administrator" [22:09:40] <klocze1> I have correctly working NIS clientn connected to this NIS server (on Linux) [22:09:49] <elektronkind> "what's your day job like, joe?" "I'm a ZFS administrator" [22:09:50] <sickness> pr0n will teach you better :P [22:09:54] <klocze1> again: NIS server is on Linux :) [22:10:24] *** mkhl has joined #opensolaris [22:10:57] *** Error_404 has joined #opensolaris [22:11:35] <klocze1> kalif: on linux it does [22:15:05] <kalif> why? [22:17:05] <kalif> I can't remember the details, but I would think that the password gets transferred in clear text over the wire [22:17:46] *** dunc_ has joined #opensolaris [22:19:58] *** yarihm has joined #OpenSolaris [22:28:49] <koolniczka> sorry for the OT: is there anyone with a hands-on experince with 50+ range users working with office applications (SOffice/OOo) on a SunRay server? [22:30:32] *** Sieghard has quit IRC [22:31:14] *** Sieghard has joined #opensolaris [22:31:44] *** aliquis has joined #opensolaris [22:34:42] *** dwc_ has quit IRC [22:38:09] *** lacaAFK is now known as laca [22:40:12] *** bengtf_ has joined #opensolaris [22:40:12] *** slowhog has joined #opensolaris [22:43:21] * Error_404 ponders for a second [22:43:29] <Error_404> why did i online this 120G drive? [22:45:07] *** mkhl has quit IRC [22:49:13] <elektronkind> because it asked you nicely [22:51:59] *** Gman has joined #opensolaris [22:52:11] <Error_404> i gues [22:52:18] <Error_404> it's got pretty terrible performance [22:52:44] <Error_404> my 160G sata pool seems to peak out at 25M bandwidth, and this drive peaks at 4M [22:56:59] *** bengtf has quit IRC [22:57:04] <jamesd_> sounds insane... my single 250GB sata drive does 65MB/s [22:59:48] <Error_404> oh, i hit 50 on the sata drive [23:00:06] <Error_404> but the PATA drive won't push past 5 [23:03:30] *** mrdeviant has quit IRC [23:07:47] *** glagasse has quit IRC [23:08:08] <hell`> jamesd_: what type of drive? seagate? [23:08:21] <jamesd_> seagate.. sun labeled. [23:08:35] <jamesd_> 7200 rpm [23:10:07] <jbalint> hi [23:11:59] <jamesd_> hello [23:12:35] <jbalint> whats up [23:18:54] *** laca is now known as lacaAFK [23:19:19] *** rookie___ has joined #opensolaris [23:19:44] <rookie___> hi all [23:20:17] <rookie___> i need some help about applying a patch on opensolaris_x86 [23:20:48] <hile_> you don't patch an OpenSolaris build [23:21:17] <rookie___> ? [23:21:23] <Tpent1> http://blogs.sun.com/tpenta/entry/non_debug_encumbered_binaries_available3 [23:21:29] <hile_> patches aren't provided for opensolaris builds [23:21:30] <Tpent1> b50 nd stuff is up [23:21:49] <hile_> how goes, alan? [23:21:57] * Tpent1 yawns [23:21:59] <Tpent1> long week [23:22:26] <hile_> i know hte feeling, and mine hasn't even started yet [23:22:30] <hile_> today's my sunday [23:22:46] *** kFuQ has quit IRC [23:23:28] *** kFuQ has joined #opensolaris [23:24:32] *** Tpenta has quit IRC [23:24:43] *** Tpent1 is now known as Tpenta [23:25:02] *** ChanServ sets mode: +o Tpenta [23:25:05] <rookie___> hmm, i think im wrong somewhere, how i can understand that my solaris is open or not? (sorry for the stupid question) [23:25:17] *** Tpent1 has joined #opensolaris [23:25:55] <rookie___> my uname -a says: SunOS solaris 5.10 Generic_118855-14 i86pc i386 i86pc [23:26:25] <Auralis> thats a standard solaris 10 [23:27:00] <Tpenta> that is in need of kernel patches [23:27:49] <rookie___> hmm, the guy who installed this system has said to me that its an opensolaris [23:28:22] <jamesd_> what does uname -av say [23:28:50] <rookie___> the same: SunOS solaris 5.10 Generic_118855-14 i86pc i386 i86pc [23:29:50] *** elflord has quit IRC [23:30:39] <rookie___> i think i need to find an solaris10 channel [23:30:41] *** triplah has joined #opensolaris [23:30:56] <Tpenta> what do you want to know abt sol 10? [23:30:57] <bengtf_> hmm, it should say something of snv... instead of generic [23:31:11] *** ds322 has joined #opensolaris [23:31:19] <rookie___> i need to apply a patch to sol 10 [23:31:35] <Tpenta> you have the patch? [23:31:36] <rookie___> i downloaded the patch, 123456 - 12.jar [23:31:41] <rookie___> yep [23:31:53] <bengtf_> hmm, funny patchnumber ... [23:32:00] <Tpenta> ok, you should extract it and use the patchadd cmd [23:32:17] <rookie___> :) patch num is an example [23:32:18] *** Error_404 has quit IRC [23:32:31] <rookie___> ok, i'll try to extract [23:32:36] *** Error_404 has joined #opensolaris [23:32:48] <Tpenta> I've never downloaded .jar files, but you can extract that with jar vxf 123456-12.jar [23:32:56] <Tpenta> then run patchadd 123456-12 [23:34:19] *** coffman has quit IRC [23:34:23] <jengelh> you can also 'unzip' it and that's it [23:34:37] <rookie___> wow! [23:34:44] <tsoome> you can also download it into /var/sadm/spool and add it with smpatch... [23:34:45] <jengelh> .jar is just a hijack of zip, and an extension theft from the JAR archiver. [23:34:48] *** Mazon is now known as mazon [23:34:56] <rookie___> its been patched [23:35:01] <jengelh> JAR being Jung Archiver, aka new-ARJ [23:35:01] <rookie___> thank you! [23:35:28] <Tpenta> :) [23:35:29] <rookie___> tssome, first i had tried with smpatch add i 123456-12 but it didnt work [23:35:48] <hile_> lesson #1: don't use smpatch :-) [23:36:20] <elektronkind> you can even "unjar" a .jar archive with the unzip commands [23:36:29] <rookie___> lesson 1: totally understood :) [23:37:16] <rookie___> anyway, its rebooting now, i'll see if the patch works or not [23:37:23] <tsoome> hm, I have been almost all patches with smpatch;) [23:37:47] <tsoome> I have used, that is:P [23:37:48] <jbalint> jengelh: Jung Archiver!? [23:39:03] <elektronkind> jbalint: if you add the right flags to unzip with dealing with a jar file, it'll tell you about the inner-meaning of your dreams [23:39:24] * jbalint `strings` it [23:39:54] <jengelh> jbalint : yep [23:40:12] <jengelh> http://www.arjsoftware.com/jar.htm [23:40:19] *** gisburn has joined #opensolaris [23:40:39] <gisburn> Where are the ipfiter error messages stored ? [23:40:46] <jbalint> Does unzip know/use this format? [23:41:01] <elektronkind> gisburn: they normally end up in /var/adm/messages [23:41:14] <elektronkind> gisburn: the ipmon process syslogs them [23:41:15] <jengelh> jbalint : Unzip knows the Java Archive (extension JAR), but does of course not know JungAR (extension J or JAR) [23:41:34] <jengelh> because java archive is just zip [23:41:50] <gisburn> elektronkind: I do not see any errors or notifications that the stuff has been reloaded [23:42:04] <jengelh> IIRC, the guy was called R.Jung [23:42:05] <jbalint> Yeah, wasn't sure if Jung was just an offshoot mention, or was actualy used in something. [23:42:17] <jbalint> jengelh: That did the ARJ stuff? [23:42:31] <jengelh> what? [23:42:32] <jbalint> Yep. [23:42:40] <jbalint> ARJ is a tool for creating compressed file archives. It was invented by Robert K. Jung. ARJ probably stands for Archiver Robert Jung. [23:42:54] <jengelh> well and JAR stands for Jung Archiver. [23:42:59] <elektronkind> gisburn: I don't think it logs /that/ sort of stuff. [23:43:08] <jbalint> Yeah, found that page on their site a few mins ago. [23:43:19] <jengelh> and has been 'invented' when Java still used .ZIP extensions [23:43:28] <gisburn> elektronkind: which is not usefull since it lacks then the important marks in the log which change may have caused trouble. [23:44:18] <gisburn> Does B48 have any known issues with accepting remote X11 connections ? [23:45:03] *** PPRO has joined #opensolaris [23:45:25] <PPRO> hi jmcp [23:45:26] <jengelh> dang wikipedia broekn [23:45:29] <PPRO> hi all [23:45:29] <jengelh> again [23:46:35] <gisburn> Which device is used by ipffilter for hme on an Ultra5 ? [23:47:07] <Tpenta> !seen stevel [23:47:08] <Drone> stevel is currently online in #opensolaris and last spoke on Thu 12 Oct 2006 15:36 GMT, saying 'morning jteo'. [23:47:24] <Tpenta> stevel: you around? [23:47:25] <rookie___> hmm, may i ask some questions more? [23:47:30] <Tpenta> always [23:47:44] <Tpenta> getting answers is usually more problematical :) [23:47:51] <elektronkind> gisburn: which device? you mean "hme0" ? [23:48:14] <rookie___> my aim is to install drivers for Sun StorEdge 2G FC PCI Dual Channel Network Adapter [23:48:24] <gisburn> elektronkind: I do not get errors or warnings, regardless whether I use hme or hme0 [23:48:29] <rookie___> so, i downloaded and applyed the patch [23:48:30] <gisburn> something is so broken [23:48:42] <elektronkind> gisburn: check to make sure you're syslogging local0 [23:48:45] <rookie___> also i can see the patch in sun update manager as "installed" [23:48:57] <elektronkind> gisburn: that's the default facility for ipmon to log to [23:49:33] <gisburn> elektronkind: how can I get messages to local0 logged ? [23:49:42] <rookie___> on the documentation it writes, after appyling the patch and rebooting you'll see "/devices/pci at b,2000/pci@2/SUNW,qlc@x dot " [23:50:04] <rookie___> but i cant see pci@b in /devices/ folder [23:50:06] <elektronkind> gisburn: add a like to /etc/syslog.conf: local0.info<tab>/file/to/log/to [23:50:10] <elektronkind> s/like/line [23:50:21] <elektronkind> gisburn: then 'pkill -1 syslogd' [23:50:38] <elektronkind> oh, 'touch /file/to/log/to' before the pkill [23:50:59] <elektronkind> syslogd won't make the file itself, it expects it to be there already [23:51:34] <gisburn> elektronkind: kill -1 == kill -HUP ? [23:51:41] <elektronkind> jah [23:51:44] <Tpenta> sorry folks, I gotta run for a bit [23:51:58] <elektronkind> jah jah, in der tjoilets [23:55:43] *** Peanut has left #opensolaris [23:55:44] *** klocze1 has quit IRC [23:56:23] <bengtf_> rookie___: did you reboot ? did it reconfigure did you install the card ? [23:56:48] <gisburn> elektronkind: this is xx at @@!! dot .. ipf does not log ANYTHING. [23:57:02] <bengtf_> also the pci@ may vary depending on ... where the card is [23:57:10] <gisburn> the whole box refuses to accept X11 connections and there is no way to enable any logs. [23:57:11] <elektronkind> gisburn: are you telling it to? [23:57:17] [23:57:20] <gisburn> xx@@@!!!! [23:57:23] <elektronkind> gisburn: heh [23:57:26] <elektronkind> gisburn: ok [23:57:41] * gisburn wants to kick someone somewhere where the sun does (usually) not shine. [23:58:04] <elektronkind> gisburn: in your ipf.conf, add a "log" directive to any line that denies packets [23:58:04] <PPRO> /etc/rcS.d/S55rdac: [ID 702911 user.error] Warning: rdriver.conf re-generation failed [23:58:16] <elektronkind> gisburn: then reload your config [23:58:35] <elektronkind> PPRO: using a A1000 are you? [23:58:45] <PPRO> nope [23:58:50] <PPRO> IBMsux DS4500 [23:59:02] <PPRO> with rdac module [23:59:05] <PPRO> sux sux sux [23:59:20] <elektronkind> I don't think that the rdac module is supported under solaris 10 [23:59:30] <PPRO> i have solaris 9 now [23:59:38] <elektronkind> ah, then it should work [23:59:47] <PPRO> should....right.. [23:59:51] <rookie___> bengtf, i was looking at the dmesg now