[00:04:27] <WickedWicky> so you didnt torch your place down then?
[00:05:28] *** Ignacio_ is now known as nachox
[00:05:41] *** furrycat has left #opensolaris
[00:06:24] *** victori_ has joined #opensolaris
[00:06:44] *** bqscott has joined #opensolaris
[00:07:59] <RainDoctor> where is sparcdr
[00:08:06] <RainDoctor> I need to know abt c90
[00:09:55] *** oninoshiko has quit IRC
[00:10:14] <bqscott> does anyone know which OpenSolaris build S10 U4 is based on?
[00:10:27] *** CIA-26 has quit IRC
[00:11:29] *** estibi has quit IRC
[00:11:52] *** moo_cow has quit IRC
[00:11:58] <nachox> none
[00:12:27] <nachox> some nevada features were backported to solaris 10 though
[00:12:52] <bqscott> I'm asking because I was hoping to see the ZFS gzip compression feature make its way into U4, and it doesn't appear to be there
[00:13:22] <bqscott> I believe that was added in b62
[00:14:40] *** asyd has joined #opensolaris
[00:21:06] *** gm135 has joined #opensolaris
[00:21:58] *** dunc__ has quit IRC
[00:22:25] *** Kernel86_ has joined #OpenSolaris
[00:23:34] *** Kernel86 has quit IRC
[00:23:53] *** JWheeler has quit IRC
[00:24:11] *** dunc__ has joined #opensolaris
[00:25:35] *** sioraiocht has joined #opensolaris
[00:26:57] *** mikefut has quit IRC
[00:29:05] *** dpn` has joined #opensolaris
[00:30:54] *** kaiwai_ has joined #opensolaris
[00:34:22] <elektronkind> uh oh
[00:34:23] <elektronkind> A security vulnerability in the X2100 and X2200 M2 Embedded Lights Out Manager (ELOM) software may allow remote unprivileged users the ability to initiate unauthorized network traffic from the embedded service processor (SP). This may allow the SP to be used as a proxy to send unsolicited bulk e-mail (spam).
[00:34:58] *** sahafeez has joined #opensolaris
[00:35:14] <WickedWicky> that's not good
[00:35:26] *** JWheeler has joined #opensolaris
[00:37:15] <bqscott> another reason to always put your management devices behind a firewall
[00:39:31] *** kaiwai__ has quit IRC
[00:39:42] <bda> Indeed.
[00:40:02] <bda> Still not as funny as people doing that with printservers, though.
[00:40:55] *** dolske has quit IRC
[00:41:25] *** estibi has joined #opensolaris
[00:41:26] *** cypromis has quit IRC
[00:42:53] *** dunc_ has joined #opensolaris
[00:44:49] *** CIA-26 has joined #opensolaris
[00:45:00] *** Kernel86_ has quit IRC
[00:45:45] *** Somethingelse has joined #opensolaris
[00:45:53] *** kimc has joined #opensolaris
[00:49:30] *** estibi has quit IRC
[00:49:41] *** estibi has joined #opensolaris
[00:52:06] *** Drone has joined #opensolaris
[00:52:08] <WickedWicky> what's the opposite of mmap?
[00:52:09] <WickedWicky> free?
[00:52:36] *** palowoda has joined #opensolaris
[00:53:11] <nachox> opposite?
[00:53:36] <WickedWicky> yes, if i understand correctly mmap() is used to allocate memory
[00:54:57] <nachox> no, it is used to map a file or shared object into memory
[00:55:04] <nachox> read the man page
[00:55:14] <WickedWicky> I am as we speak
[00:55:39] <WickedWicky> munmap
[00:55:44] <WickedWicky> that's what I was looking for
[00:56:25] <dlg> you can do anon mmaps and use it as allocated memory
[00:56:51] <WickedWicky> I am trussing something, and i see it do 8043984398439843 mmaps
[00:57:03] <WickedWicky> so I wanted to grep the logfile to see if it was "unmapping" as well
[00:57:28] <WickedWicky> since i see my memory usage go up with the second
[00:57:28] <dlg> you're right, munmap is wht you wanted
[00:58:11] <WickedWicky> one day I'll learn how to proper debug ;-)
[00:58:22] *** dunc__ has quit IRC
[00:58:58] *** jteo has quit IRC
[01:04:08] *** richlowe has joined #opensolaris
[01:06:38] *** gm135 has quit IRC
[01:13:28] *** dolske has joined #opensolaris
[01:16:12] *** estibi has quit IRC
[01:18:23] *** SirFunk has joined #opensolaris
[01:30:28] *** delewis has quit IRC
[01:31:43] *** cypromis has joined #opensolaris
[01:31:53] <sahafeez> anyone - booting a cdrom on a sun (ide) that does not have the cdrom set right - boot /pci1f,0/ide@d/cdrom@1,0;f ? the cdrom is 2nd master
[01:31:58] *** delewis has joined #opensolaris
[01:34:34] *** Fullmoon has joined #opensolaris
[01:36:02] *** perlmonk has quit IRC
[01:36:18] *** andy_ has joined #opensolaris
[01:38:21] *** Andy_Pease has quit IRC
[01:40:25] *** nachox has quit IRC
[01:40:27] *** kFuQ has quit IRC
[01:42:00] *** bunker has quit IRC
[01:42:54] *** blindfish has quit IRC
[01:52:55] *** CIA-26 has quit IRC
[01:54:37] *** dlg has quit IRC
[02:08:18] *** peemus-home has joined #opensolaris
[02:14:57] *** ]ramzi[ has joined #opensolaris
[02:17:54] *** furrycat has joined #opensolaris
[02:18:27] <furrycat> i would like to recommend that we rename the packages
[02:18:48] <furrycat> it should be JAVAemacs rather than SUNWemacs for example
[02:19:08] <furrycat> this way we are consistent with the stock ticker.
[02:19:14] <furrycat> it helps branding
[02:19:45] <kjetilho> they already made an exception to that rule when they renamed SUNWspro SPROfoo
[02:22:58] <jamesd> furrycat, jonathan allready said they are not renaming packages...
[02:23:18] <furrycat> Jonathan needs to now that he renamed the stock ticker!
[02:23:19] <jamesd> besides associating emacs with java, might make  java seem less bloated...
[02:23:58] <furrycat> I have a JDS and use SMC, and my packages should be JAVAsmc for example
[02:24:40] <jamesd> are you trying to be the sacrificial  saturday night troll?
[02:24:45] *** peemus-home has quit IRC
[02:24:46] <furrycat> yeah
[02:24:54] <Doc> hey, that's my job!
[02:24:58] <furrycat> one too many.  CUI
[02:25:04] <richlowe> Doc: you limit yourself to saturdays?
[02:25:24] <Doc> i'm trying to cut back - i only do saturday, tuesday and thursday now
[02:25:28] <furrycat> it would be totally funny if some marketing droid actually wanted to rename the packages.
[02:25:29] *** peemus-home has joined #opensolaris
[02:25:30] <richlowe> trying the same troll twice was totally pushing it though.
[02:26:10] <peemus-home> has anyone ever had the following error when booting sxde, or s10u4?
[02:26:11] <peemus-home> ata_command: select failed DRDY 0x1 CMD 0xef F 0x3 N 0x0  S 0x0 H 0x0 CL 0x0 CH 0x0
[02:26:35] <peemus-home> sxce b73 seems to boot it fine though
[02:27:16] <palowoda> The workaround is to use sxce.
[02:27:30] <peemus-home> :)
[02:27:40] *** furrycat has left #opensolaris
[02:27:45] *** jamesd has quit IRC
[02:29:56] *** CIA-17 has joined #opensolaris
[02:30:25] *** furrycat has joined #opensolaris
[02:30:33] *** jamesd has joined #opensolaris
[02:30:34] *** ChanServ sets mode: +o jamesd
[02:30:38] <furrycat> can someone sanitize the topic for #solaris
[02:33:48] <jamesd> no.. its not the type of channel it is...  we have rough edges and we let them show...
[02:34:11] *** mmagnon has quit IRC
[02:35:47] <bda> Rough edges. Is that the phrase? :)
[02:37:12] *** mcnamarabrian has joined #opensolaris
[02:37:40] *** mcnamarabrian has quit IRC
[02:41:28] *** bqscott has left #opensolaris
[02:42:38] *** paulf has joined #opensolaris
[02:43:28] <furrycat> jamesd: will it be fixed in the next update?
[02:44:36] *** furrycat has quit IRC
[02:45:14] *** yongsun has joined #opensolaris
[02:56:11] *** cypromis_ has joined #opensolaris
[02:59:16] *** NikolaVeber has quit IRC
[02:59:40] *** mcnamarabrian has joined #opensolaris
[03:00:46] *** TheZeroX has quit IRC
[03:00:47] *** dunc_ is now known as dunc
[03:01:39] *** dolske has quit IRC
[03:06:19] *** cypromis has quit IRC
[03:09:22] *** andy_ has quit IRC
[03:13:42] *** Auralis_ has joined #opensolaris
[03:13:47] *** crib has quit IRC
[03:14:08] *** Auralis has quit IRC
[03:14:53] *** crib has joined #opensolaris
[03:15:49] *** masta has joined #opensolaris
[03:16:14] *** dpn` has quit IRC
[03:16:59] *** bondolo has quit IRC
[03:17:38] *** dpn` has joined #opensolaris
[03:17:59] *** mcnamarabrian has quit IRC
[03:20:06] *** Mdx4 has joined #opensolaris
[03:22:25] *** Fullmoon has quit IRC
[03:22:42] *** GmanAFK has quit IRC
[03:24:03] *** Pietro_S has quit IRC
[03:25:06] *** Odin-SOL has joined #opensolaris
[03:26:29] *** Syndelor has joined #OpenSolaris
[03:27:31] <Syndelor> Evening all
[03:27:49] <Odin-SOL> Evenin'.
[03:27:58] *** Disreali has quit IRC
[03:29:27] <paulf> Hello
[03:29:52] <jbk> hi
[03:29:59] *** ]ramzi[ has quit IRC
[03:31:22] <jamesd> hola
[03:31:45] <Syndelor> Replaced to K class 9000's today with two dl 365's Solaris 10 nothing like turnign support costs from 20 thousand a year to 2 lol
[03:32:13] *** perlmonk has joined #opensolaris
[03:34:28] <RainDoctor> hi guys
[03:34:45] <RainDoctor> where can I get a list of supported graphics cards, video cards, wireless cards?
[03:34:48] <RainDoctor> for opensolaris
[03:35:03] <jbk> Syndelor: i bet your cooling costs will drop dramatically too
[03:35:42] <jamesd> RainDoctor,   sun.com/bigadmin/hcl for a start...
[03:35:53] <jbk> those K boxes put out a LOT of heat
[03:36:42] <Syndelor> No shit I will be able to kill an ac unit in the small dc that the bxes where in as soon as I get hte old symetric turned off as well
[03:37:00] <Syndelor> They are in an airport closet not really a dc lol
[03:37:15] *** Gman has joined #opensolaris
[03:38:00] *** Odin-SOL has left #opensolaris
[03:39:53] <RainDoctor> do all intel wireless chipsets work on opensolaris?
[03:39:57] <RainDoctor> sorry for bothering
[03:40:06] <RainDoctor> I couldn't find the one on that swebsite
[03:40:13] <RainDoctor>   http://www.shopping.hp.com/webapp/shopping/computer_can_series.do?storeName=computer_store&category=notebooks&series_name=dv9500t_series&tab_switch=true&catLevel=3&tab=specs
[03:40:36] <RainDoctor> I am looking at this laptop to see whether I can install opensolaris there
[03:41:22] <kaiwai_> nope, the wireless on that laptop isn't supported yet
[03:41:27] <RainDoctor> thanks
[03:41:38] <jamesd> sorry i dont do  wireless  on solaris
[03:41:47] <Syndelor> What you have
[03:41:59] <kaiwai_> RainDoctor: a good laptop I would suggest is the acer gemstone which includes an atheros wireless + Nvidia graphics etc.
[03:42:12] <RainDoctor> kaiwai_,let me check that laptop
[03:42:17] <Syndelor> I ma runnign on an nw9440 with the latest intel wireless and it works awsome on Solaris-Dev ed version 64a but not on 66
[03:42:42] <kaiwai_> IIRC you can pick up the laptop for around NZ$1400
[03:42:44] <Syndelor> It is all in the driver set shipped
[03:43:16] <kaiwai_> Symmetria: whats the wireless chipset?
[03:43:17] <Syndelor> I wouldnt suggest an acer to anyone... My wife has one it is horrible
[03:43:27] <Syndelor> Noisy and HOT
[03:43:34] <RainDoctor> Syndelor, nw9440 hp?
[03:43:57] <RainDoctor> I dont wanna shell out US $1800 on a laptop
[03:43:58] <kaiwai_> IIRC dell sell ones with the 3945 wireless but the downside the only one that supports WPA is atheros wireless
[03:44:26] <Syndelor> Yeah it is the mobile workstation with NVIDIA vid card (AWSOME FOR UNIX ANYTHING) and the intel wireless card and broadcom wire card only thing not working is the sound card and I beleive that is because I have not taken the time to fix it
[03:44:27] <jbk> still? I thought at theast the intel chips would support wpa by now
[03:44:45] <Gman> jbk, soon, soon - or so we're all told :)
[03:44:50] <Syndelor> Not saying you should just compare the hardware
[03:45:28] <Syndelor> I use wep everywhere wpa is banned onsite with the cisco agent requierment except for my mac
[03:45:31] <kaiwai_> jbk: when Windows Vista is fixed, so will wpa will arrive for Intel wireless on solaris :)
[03:46:15] <RainDoctor> hmm nw9440 is $2799
[03:46:22] <RainDoctor> I cant afford that much
[03:46:35] <kaiwai_> yeap, the one I've got, dv6209tx works nicely
[03:46:44] <kaiwai_> that was NZ$2499 incl GST
[03:46:45] <jbk> i'm going to feel so helpless without my laptop here for the next few weeks...
[03:46:50] *** ballchalk has joined #opensolaris
[03:48:36] *** simford has joined #opensolaris
[03:50:40] <RainDoctor> hmm
[03:50:47] <ballchalk> hey guys does tkip auth wifi work with intel 2200ng?
[03:50:51] <ballchalk> Bg
[03:50:52] <RainDoctor> no hope for $1000 laptops
[03:50:52] <kaiwai_> nope
[03:51:13] <kaiwai_> Atheros is the only one that supports WPA
[03:51:15] <Syndelor> I wasnt saying get one of them just compare hte hardware
[03:51:26] <ballchalk> Sh1t
[03:52:08] <ballchalk> Kaiwai thanks
[03:52:20] *** ballchalk has quit IRC
[03:52:44] <kaiwai_> ballchalk: I feel your pain; I want wpa for intel 3945 a/b/g but there isn't alot I can do about it
[03:52:58] *** cypromis_ has quit IRC
[03:54:07] *** kaiwai_ is now known as kaiwai
[03:54:53] <Syndelor> Anyone here a MAC fan
[03:55:31] <Syndelor> The new Powerbook is great for Solaris but it is extremely pricy
[03:56:40] <kaiwai> Syndelor: hmm, yeah, but you'd have to deal with EFI and hacking solaris
[03:56:53] <kaiwai> Toshiba are supported well
[03:58:22] *** simford has quit IRC
[04:02:15] *** newpers has left #opensolaris
[04:03:28] *** Mdx4 has quit IRC
[04:03:39] <RainDoctor> this hp laptop comes with two hard drives, which is a way-cool
[04:04:58] *** Kernel86 has joined #OpenSolaris
[04:05:49] <Syndelor> Yeah I have two in mine one is a slide in that takes the place of the cdrom
[04:05:53] <Syndelor> Keep that in mind
[04:06:26] <Syndelor> Not with the new core 2 duo systems it worked from boot including sound and wifi and hardline nic
[04:06:47] <RainDoctor> hmm, this one not in the place of cdrom
[04:07:06] <RainDoctor> it is 17 inches laptop
[04:07:09] <RainDoctor> weighs 8 lbs
[04:07:13] <Syndelor> Yeah it is man it is a replacement drive for CDrom I can almost guarentee it or it is external
[04:08:09] <RainDoctor> oh well, it comes with cdrom and with 2 hard drives
[04:08:14] <RainDoctor>   http://www.shopping.hp.com/webapp/shopping/computer_can_series.do?storeName=computer_store&category=notebooks&series_name=dv9500t_series&tab_switch=true&catLevel=3&tab=specs
[04:09:27] <Syndelor> Yeah the cdrom and hardrive kit are hotswapable
[04:09:38] <Syndelor> We use a model similar to that at work
[04:10:47] <RainDoctor> ok
[04:12:06] <kaiwai> hmm, 8lbs, that would be around 4kg IIRC
[04:12:14] <RainDoctor> yeah
[04:12:17] <RainDoctor> 8/2.2
[04:12:20] <kaiwai> (for those of us living in the 21st century)
[04:12:21] <jbk> a little under
[04:14:08] *** Odin- has quit IRC
[04:16:13] *** victori_ has quit IRC
[04:32:41] *** Triskelios has quit IRC
[04:34:57] *** yongsun has quit IRC
[04:36:42] *** victori_ has joined #opensolaris
[04:39:29] *** CIA-17 has quit IRC
[04:45:19] *** kaiwai has quit IRC
[05:01:14] *** Gekz[sleep] is now known as Gekkko
[05:01:16] *** Gekkko is now known as Gekz
[05:01:36] *** paulf has quit IRC
[05:04:24] *** infidel has joined #opensolaris
[05:08:33] *** Netwolf_ has joined #opensolaris
[05:09:35] <infidel> Hi, I just installed OpenSolaris on my PC and I am having trouble getting it to talk to my Windows laptop
[05:13:29] *** bengtf has quit IRC
[05:13:46] <infidel> does anyone know what might cause OpenSolaris to not reply to ARP requests?
[05:14:15] <Tempt> That's very odd.
[05:14:34] <Tempt> Use snoop on your OpenSol box to watch the network interface.
[05:15:02] <infidel> ok thanks I will try it
[05:20:51] *** Netwolf has quit IRC
[05:29:16] *** Chihan has quit IRC
[05:30:04] <infidel> when my laptop sends an ARP request to my Solaris PC, it is not visible in the snoop output
[05:31:00] <infidel> but the Solaris PC can access the internet via my DSL router
[05:31:13] <Tempt> Well
[05:31:18] <Tempt> perhaps you've got some crazy vlan setup.
[05:31:23] <Syndelor> You on wifi
[05:31:30] <infidel> wifi yes
[05:31:39] <Syndelor> You on the right one
[05:31:52] <infidel> yes
[05:32:27] <Syndelor> go to ipchicken on both machines make sure you get same IP
[05:32:30] <infidel> on my laptop, I can see some traffic from the Solaris PC using WireShark
[05:32:53] <Syndelor> SO can you shell into it from the windows box
[05:32:58] *** bengtf has joined #opensolaris
[05:33:37] <infidel> no, I can't talk to it
[05:34:38] <infidel> it looks like solaris is responding to ARP requests from my router, and it is also sending gratuitous ARP sometimes
[05:37:35] <Tempt> Wierd.
[05:39:01] <infidel> on snoop I sometimes see "Ether Type=8888E (Unknown)". does that mean anything?
[05:43:59] *** Shiv_1 has joined #opensolaris
[05:45:12] *** Marv|LG has quit IRC
[05:45:39] *** Marv|LG has joined #opensolaris
[05:47:54] *** Shiv_1 has quit IRC
[05:48:28] *** Shiv_1 has joined #opensolaris
[05:52:53] <libkeiser> that's strange output. i thought the ethertype field was only 16bits...
[05:54:18] <infidel> sorry typo, 888e is apparently "IEEE Std 802.1X - Port-based network access control"
[06:02:48] <Tempt> WPA
[06:04:03] <infidel> but the fact that it says "unknown" is just a limitation of snoop, it doesn't mean that there is a problem with WPA, right?
[06:04:27] <Tempt> Nah, but some sort of wierdo WPA problem might lead to this.
[06:04:34] <Tempt> Man, wifi, it drives me mad.
[06:05:28] <Tempt> WickedWicky: PING
[06:05:37] <Tempt> WickedWicky: Pebbles appears to have suffered a mysql death.
[06:08:44] <infidel> yeah wifi is annoying when there are problems, but it is so useful and most of the time it works for me so I don't have to think about it
[06:09:12] <sahafeez> someone here have a good understanding of obp and ide devices?
[06:09:26] <delewis> there's cheap wifi *cough* Linksys *cough* and there's real wifi -- Cisco.
[06:09:33] <delewis> and don't say they're the same.
[06:10:10] <sahafeez> no, linksys never said you could fill all the ports and run it at wire speed without it crashing. cisco tells you you can :)
[06:11:22] <delewis> I can't wait until I replace this Ultra 2 with a v480, x4100, or x4150.
[06:11:54] <delewis> ZFS compression just kills it.
[06:12:12] <delewis> and it only supports up to 2GB of memory.
[06:12:14] <delewis> :-/
[06:12:55] *** oninoshiko has joined #opensolaris
[06:12:57] <Tempt> Hey, don't be knocking the finest workstation Sun ever made.
[06:13:33] <delewis> I wouldn't call it the 'finest', but it certainly isn't a terrible workstation.
[06:13:40] <Tempt> Finest.
[06:13:41] <Tempt> Easily.
[06:13:50] <Tempt> When you factor in what the competition was doing.
[06:13:52] <delewis> I'd have trouble calling anything the 'finest' from the defunc UltraSPARC-II generation.
[06:14:03] <delewis> defunct, rather.
[06:14:41] <Tempt> A lot of people believe a number of US car manufacturers had their finest moment many, many years ago.
[06:15:07] <delewis> the finest workstation Sun has ever made is the Blade 1000/2000, IMO.
[06:15:21] <Tempt> A truly nice workstation, yes.
[06:15:24] <delewis> none of the competition had features that it had.
[06:15:33] <Tempt> But not as exciting and revolutionary as the Ultra-2.
[06:15:41] <delewis> the Ultra 2 was a run of the par workstation by the competition standards, which included the HP9000 J2240 and the SGI Octane.
[06:15:52] <delewis> Tempt: bullshit. the Ultra 2 didn't have on-board fibre.
[06:16:29] <Tempt> To be honest, I'm not sure onboard FCAL is that much of a winning feature. It certainly shipped with SBus HBAs as an option.
[06:16:30] <delewis> the Ultra 2 was also limited IO-wise, because of SBus. At the time SGI had XIO in the Octane 2, which was capable of 3.2GBps.
[06:16:52] <Tempt> Remember, 64 bit SBus had more bus bandwidth than PCI.
[06:17:10] <delewis> Tempt: not PCI-E, which is what the J2240 had. and SGI was using XIO.
[06:17:11] <Tempt> You could also probably buy four Ultra-2s for the price of a loaded up Octane 2
[06:17:23] <jamesd> delewis, not really limited at the time.. no one needed more than 100MB/s    and even that was excessive
[06:17:34] <Tempt> PCI-E wasn't even a standard when these machines were released.
[06:17:40] <jamesd> the bus was  2GB/s  for the video card...
[06:17:48] <delewis> Octane2 wasn't even available when the Ultra 2 was around. a base Octane config sold for $20,000-$30,000, which is what a decent Ultra 2 config would've cost.
[06:18:15] <delewis> Tempt: I did have a J2240 from 1997 that had 64-bit PCI.
[06:18:27] <Tempt> delewis: 64bit PCI != PCI-E
[06:18:45] <delewis> is 64-bit PCI PCI-X?
[06:18:49] <delewis> I always get them confused.
[06:18:57] <Tempt> I think PCI-X is the current term, yes.
[06:19:03] <delewis> ok, that's what I meant then.
[06:19:09] <Tempt> I hate this silly naming scheme, I just name 'em by width and speed.
[06:19:14] <delewis> yeah.
[06:19:14] <jamesd> pci-x is  100/133mhz  64 bit pci.
[06:19:35] <Tempt> Okay, I think delewis means 64bit PCI, probably 66Mhz.
[06:19:36] <delewis> really? I thought it also included 33/66MHz 64-bit PCI, as well.
[06:19:49] <Tempt> Oh, you'll love it, they've got magical names for it all now
[06:19:54] <delewis> fuck.
[06:19:57] <Tempt> PCI-1X, 2X, 8X ...
[06:20:03] <jamesd> it can down grade...  but it wasn't really pci-x with out the 133mhz part.
[06:20:04] <Tempt> none of which is particular descriptive or useful.
[06:20:09] <Tempt> Marketdroid enhanced
[06:20:11] * delewis makes it a point to not familiarize himself with peecee terminology
[06:20:27] <Tempt> You see, the PC people used AGP with a multiplier
[06:20:38] <Tempt> so marketdroids figured PCI should have a similar naming scheme
[06:20:53] <delewis> that's what I liked about the simplicity of UPA. It scalled with the system bus speed.
[06:21:09] <Tempt> Fireplane, man, fireplane.
[06:21:10] <delewis> ther was no UPAxN (where N = 1, 2, 3, 4, etc.)
[06:21:17] <delewis> there*
[06:22:34] <Tempt> UPA iss nice and all, but Gigaplane and Fireplane are much more interesting.
[06:22:44] *** jte1 has joined #opensolaris
[06:22:58] *** jte1 is now known as jteo
[06:23:01] <delewis> Tempt: UPA co-exists with Gigaplane and Fireplane.
[06:23:13] <Tempt> Hmm
[06:23:17] <Tempt> Which machines mixes UPA and Fireplane?
[06:23:20] <delewis> UPA is the graphics-memory interconnect, whereas Gigaplane and Fireplane are the system interconnects.
[06:23:30] <delewis> Tempt: any UltraSPARC-III generation system.
[06:23:35] <delewis> the Blade 1000, for example.
[06:23:43] <Tempt> aaah
[06:23:45] <Tempt> of course
[06:24:45] <Tempt> UPA was such a nice bus, perfect for the job years before the pc bucket brigade worked out anything similar.
[06:24:55] <Tempt> Pity Sun never woke up to the concept of blazing fast framebuffers.
[06:25:23] <delewis> yep, UPA was a much more elegant solution to PCI and AGP, which are suitable for exclusive workloads (programmed v. texture)
[06:25:45] <jamesd> they decided they didn't want to go head to head with SGI, they decided to do the fileserver part instead.
[06:25:48] <delewis> and like I said, it just scaled with the system bus speed.
[06:25:49] <Tempt> Still, I'm waiting to try a system with graphics on the fireplane.
[06:26:01] <delewis> Tempt: XVR-4000?
[06:26:06] <Tempt> Yes.
[06:26:16] *** sahafeez has quit IRC
[06:26:19] <delewis> because the XVR-4000 is sitting on the CPU interconnect, I would think it'd be on the Fireplane.
[06:26:26] <Tempt> Although I don't know if I'd want to sacrifice the CPU slots for it.
[06:26:59] <Tempt> If I could get my fingers on an zulu I'd definately give it a go and see how it performs.
[06:27:04] <Tempt> I'd expect it to be pretty snappy.
[06:27:15] <delewis> Fireplane only connects two processors to the memory controller, though, right?
[06:27:30] <delewis> so only the processor that's on the same interconnect as the XVR-4000 would be able to fully utilize it.
[06:27:32] <Tempt> Fireplane is the whole backplane interconnect.
[06:27:50] <Tempt> In the daktari, all the CPUs and XVR-4000s are on the fireplane.
[06:28:10] <delewis> Tempt: that's not how it works from my understanding. For example, the Blade 1000 has two Fireplanes for each processor and each of those as direct access to the memory controller.
[06:28:13] <Tempt> (the I/O planar also has a fireplane connection, but only to connect bridges)
[06:28:48] <Tempt> I'd have to grab my 880 doco to double-check, but my understanding is there is a fireplane interconnect between all four SBs in the daktari
[06:28:59] <delewis> hmm, OK.
[06:29:08] <Tempt> It involves a lot of crossbars.
[06:29:13] <delewis> right
[06:29:36] <delewis> my understanding is that at most a Fireplane interconnects two processors and there are slower crossbars that interconnect those.
[06:29:53] <Tempt> Again, I'd have to check up.
[06:30:06] <Tempt> Perhaps in the spirit of marketing-speak they just call everything a fireplane.
[06:30:13] <delewis> much as how the CPU interconnects on the E4500 and friends were faster than the crossbar that interconnected each system board.
[06:30:47] <delewis> it was all Gigaplane but Sun made a distinction between the Gigaplane on the boards and the Gigaplane that interconnected each board.
[06:30:53] <Tempt> Mmm.
[06:31:22] <delewis> (3.2GBps v. 1.6GBps if memory serves me correctly)
[06:31:40] *** alanc_away_ is now known as alanc_away
[06:34:56] *** ashner has joined #opensolaris
[06:35:08] * Tempt grabs the 880 manual from the shelf.
[06:36:14] <Tempt> They definately refer to the centreplane connectivity as FirePlane
[06:36:40] <Tempt> and claim 9.6Gbps
[06:36:40] <delewis> sure, but the centerplane could just be a processor-memory-IO interconnect.
[06:36:55] <delewis> the question is how many processors are on each of those interconnects
[06:36:59] <Tempt> There's lots of pretty block diagrams
[06:37:04] <Tempt> :)
[06:37:06] *** perlmonk has quit IRC
[06:38:46] <Tempt> Okay, the two CPUs and memory on an SB has a stack of connectivity on a local crossbar - five ports, two cpu, two memory and one backplane
[06:38:56] <Tempt> each is a seperate fireplane bus.
[06:39:03] <delewis> that's what I thought
[06:39:08] <delewis> 2 procs per interconnect
[06:39:17] <Tempt> Yep.
[06:40:17] <delewis> so only the processor on the same interconnect as the XVR-4000 could fully utilize it, as that that interconnect directly connects the processor, the XVR-4000, and the memory to pull textures out of.
[06:40:31] <Tempt> There's some entertaining reading in the training manuals.
[06:40:32] <delewis> otherwise, you'll be going over the crossbar and seeing a hit on throughput.
[06:40:54] <Tempt> Well, all communication between the XVR and the CPU would be over the crossbar.
[06:41:22] <delewis> that shouldn't be the case with the processor that's sharing the interconnect with the XVR-4000.
[06:41:28] <delewis> XVR-4000 only takes up one CPU slot, right?
[06:41:38] <Tempt> Yes, but each slot accomodates two CPUs.
[06:41:44] <delewis> oh, gotcha
[06:41:46] <Tempt> There are four system board slots, each one caters to two CPUs.
[06:41:51] <delewis> right
[06:41:51] <Tempt> the XVR takes a whole SB slot
[06:41:55] <Tempt> (physically massive!)
[06:42:07] <delewis> in that case, you're right about any IO with the XVR-4000 going over the crossbar.
[06:42:10] * delewis wasn't thinking
[06:44:01] <delewis> the XVR-4000 still has direct access to a memory controller, though :-)
[06:45:20] <Tempt> The XVR-4000 was a nice idea, though. Pity it was so short-lived.
[06:46:34] <delewis> an expensive idea, though. A base v880 was what? $40,000? XVR-4000 was probably $30,000 or so, itself, so we're talking $70,000 for a graphics solution.
[06:47:30] <Tempt> True, true.
[06:47:49] <Tempt> And in all honesty, most likely not that an impressive a solution compared to other vendors offerings of the day.
[06:48:36] <delewis> at the low-end, you had Nvidia and 3dlabs, and the high-end was Infinite Reality.
[06:48:48] *** dlynes_ has quit IRC
[06:49:24] <Tempt> Pretty much.
[06:49:34] <Tempt> And Sun was having a bad time with graphics then
[06:49:40] <delewis> a decent 3dlabs-based visualization system was probably $10,000-$20,000 and IR2 was $200,000 minimum (Oynx2)
[06:49:54] <delewis> and IR2 is awesome (and still is)
[06:49:58] <Tempt> For the price of an XVR-500, you could buy a case of Nvidia cards with better real-world performance.
[06:50:00] <delewis> s/is/was/
[06:50:36] <jteo> exotic hardware has rarely been value for money.
[06:50:56] <delewis> IR2 still has a hold.
[06:51:10] <delewis> it's mostly used for visualization systems that require a lot of screen real-estate.
[06:52:21] <delewis> it goes up to 133 million pixels per system, IIRC.
[06:52:36] <delewis> current Nvidia cards (even Quadros) don't get *anywhere* near that.
[06:53:03] <Tempt> Mmm
[06:53:29] <delewis> newer quadros do something like 8-16 million pixels.
[06:53:36] <Tempt> Sun just hasn't a good run with workstation graphics.
[06:53:43] <the-decider> you could also flexibly configure the output; e.g. parts of your graphics pipe could be going out on this format, other parts in a nother
[06:53:46] <delewis> and IR2 is roughly 5-7 years old.
[06:54:07] *** dlynes_laptop has joined #opensolaris
[06:54:31] <delewis> sure, it's expensive, but when you need to push visuals out to huge displays and lots of them, it's about the only solution.
[06:54:47] <delewis> and the XVR-4000 was in direct competition with it.
[06:55:20] <Tempt> And the XVR-4000 lost. It couldn't push that sort of pixel count on a good day with optimal workloads, let alone in a real world situation.
[06:56:26] <the-decider> kinda sucks that there's nobody around pushing the envelope on a grand scale w/ the graphics stuff anymore
[06:56:55] <delewis> yep, and I'm not exactly sure what IR2's path is with SGI's current financial situation.
[06:57:43] <delewis> I'm sure it'll be that and the Altix stuff which keeps SGI afloat in the end, though, as IR2 is the only graphics solution at the moment that offers that kind of throughput.
[06:57:56] <the-decider> a bunch of pcs with nvidia cards, connected with infiniband, and with a whole lot of extra coding to distribute your graphics output plus a bunch of edge-aligned projectors ;)
[06:58:24] <delewis> the-decider: sure, but that's a very custom job that'd probably cost enough FTE hours to justify the cost of an IR2 system.
[06:58:35] <delewis> that's been tried and tested and works. :-P
[06:58:38] <the-decider> yup :(
[06:59:11] *** Gman has quit IRC
[06:59:29] <Tempt> You'd be surprised
[06:59:39] <Tempt> a bar down the road has a wrap around video setup
[06:59:44] <Tempt> with 8 edge aligned projectors
[06:59:51] <jteo> why buy when we can reinvent the wheel?
[07:00:14] <delewis> and re-solve all the problems SGI faced with the R&D work on the IR2. :-)
[07:00:35] <the-decider> Tempt: yeah, but if you're trying to do a truly "realistic" flight sim, and need to render in real time one coherent 3d model space, something like the IR2 makes it easy
[07:00:40] <delewis> Tempt: 8 projectors running at a high-resolution by today's standards is still a far cry from the 133 million pixels an IR2 can put out.
[07:00:57] <Tempt> Oh, sure, but they built this out of commodity hardware for a bag of peanuts
[07:00:58] <the-decider> splitting up raster-images and doing simple effects for video walls is arther easy.
[07:01:14] <Tempt> and a lot of those extremely sexy graphics solutions end up doing mundane jobs
[07:01:21] <delewis> yep, and the IR2 was capable of doing rendering and texture mapping on those 133 million pixels, too. :-P
[07:01:27] <Tempt> like fancy status displays in NOCs
[07:01:38] <delewis> Tempt: or ATC.
[07:01:51] <delewis> Matrox has that market cornered nowadays, though.
[07:01:56] <Tempt> Which doesn't require any real 3D performance, just huge pixel count.
[07:02:20] <delewis> the market for IR2 seems to be integration with clusters *cough* Altix *cough* to visualize data.
[07:02:39] <delewis> LRNL has one IR2 setup to visualize nuclear simulations.
[07:03:00] <delewis> ORNL has one, as well.
[07:03:03] <Tempt> Areas with traditionally massive budgets.
[07:03:35] <the-decider> aka your tax dollars at work ;)
[07:04:11] <delewis> can't be any more of a waste of money than using it to build secret prisons in eastern europe :-)
[07:05:29] * Tempt is not a US citizen.
[07:05:48] <the-decider> ..ok, well, at least *my* tax dollars at work
[07:05:58] <Tempt> The Australian government doesn't believe in heavy scientific spending like that.
[07:06:01] <delewis> you're still subject to our policies, unfortunately, though. :-(
[07:06:28] <delewis> those eastern european countries had to give approval in some form or the other (even if it was at gun point) for us to build those prisons.
[07:06:28] <Tempt> of course
[07:06:39] <delewis> and likewise we have HAARP arrays setup in Alaska, but also in Norway.
[07:06:46] <Tempt> That's part of being a tin-pot country led by a wanker with GWB's cock down his throat.
[07:06:54] <Tempt> Might change at the next election though.
[07:07:23] <palowoda> Darn there goes RDI tadpole sales.
[07:07:27] *** charlesnw has quit IRC
[07:08:41] <delewis> http://maps.google.com/maps?q=Rachel,+Nevada&ll=37.401237,-116.236038&spn=0.024376,0.042658&t=k&hl=en
[07:08:52] <delewis> at the moment we seem to be fascinated with building concentric circles.
[07:09:09] <delewis> (that's within the NTS -- Nevada Test Range, which is our nuclear test grounds)
[07:10:15] <the-decider> those are ufo landing spots ;)
[07:10:44] <delewis> http://maps.google.com/maps?q=Rachel,+Nevada&ll=37.663536,-116.024294&spn=0.062399,0.084801&t=k&hl=en
[07:10:48] <delewis> there's another one
[07:11:01] <delewis> (even spookier, because it's contained within a triangle)
[07:13:12] <Tempt> Quick! Mulder! Scully!
[07:13:19] <delewis> :-)
[07:13:54] <bda> Someone been reading some Charles Stross? ;p
[07:17:50] *** masta has quit IRC
[07:21:20] *** CIA-26 has joined #opensolaris
[07:34:26] *** nostoi has joined #opensolaris
[07:34:40] *** estibi has joined #opensolaris
[07:37:06] *** oninoshiko has quit IRC
[07:58:42] *** infidel has quit IRC
[07:58:51] *** derchris^ has quit IRC
[07:59:06] *** derchris^ has joined #opensolaris
[08:07:44] *** gavagai_ has joined #opensolaris
[08:09:59] *** simford has joined #opensolaris
[08:18:20] *** m0le has quit IRC
[08:22:24] *** sahafeez has joined #opensolaris
[08:25:32] *** RainDoctor has quit IRC
[08:26:42] *** manish has joined #opensolaris
[08:33:00] *** loke has joined #opensolaris
[08:36:45] *** estibi_ has joined #opensolaris
[08:37:13] *** RainDoctor has joined #opensolaris
[08:44:24] *** estibi__ has joined #opensolaris
[08:45:00] *** manish is now known as monzie
[08:46:42] *** Shiv__ has joined #opensolaris
[08:53:42] *** gavagai_ has quit IRC
[08:55:04] *** mschenck has quit IRC
[08:56:34] *** mschenck has joined #opensolaris
[08:59:00] *** gavagai_ has joined #opensolaris
[09:00:07] *** estibi has quit IRC
[09:04:42] *** estibi_ has quit IRC
[09:10:39] *** yongsun has joined #opensolaris
[09:13:33] *** estibi__ is now known as estibi
[09:14:14] *** Shiv_ has joined #opensolaris
[09:16:37] *** RainDoctor has quit IRC
[09:20:22] *** Shiv_ has quit IRC
[09:20:39] *** Shiv_ has joined #opensolaris
[09:21:05] *** Shiv_ has quit IRC
[09:27:51] <e^ipi> heh, I made my own "keep coffee fresh longer" valve bag out of an old pound of *$ coffee and a ziplock
[09:30:18] *** Syndelor has quit IRC
[09:30:35] *** Shiv_1 has quit IRC
[09:31:00] *** Shiv__ has quit IRC
[09:46:38] <Tempt> e^ipi: heh
[09:46:51] <Tempt> e^ipi: probably easier just to consume all your coffee within a week of roasting ;)
[09:47:16] <trochej> Elo
[09:47:26] <Tempt> 'lo.
[09:47:38] <Tempt> or perhaps even.. EHLO
[09:48:10] <trochej> :)
[09:48:16] <trochej> I had
[09:48:28] <jteo> i prefer an IV drip of caffeine.
[09:48:29] <trochej> On sunday and the last day of my work here
[09:48:30] <e^ipi> Tempt: or not roast batches big enough for it to matter
[09:48:40] <trochej> To break the server :/
[09:48:54] <Tempt> heh
[09:48:57] <Tempt> destruction
[09:49:17] <e^ipi> http://img410.imageshack.us/my.php?image=dscf0384ch1.jpg <-- mailman brought me presents yesterday
[09:49:48] *** peemus-home has quit IRC
[09:50:05] <WickedWicky> morning all
[09:50:32] <Tempt> e^ipi: Ooh, what did you get?
[09:50:55] <e^ipi> an african sample pack
[09:51:26] <e^ipi> 1/4 lb each of yrgacheffe, tanzanian peaberry, malawi & burundi
[09:52:26] <e^ipi> I obliterated half the bag of burundi with a frypan & when that turned out to be a miserable failure I went to the thrift store & got a popcorn popper to roast the other half
[09:52:35] *** monzie has quit IRC
[09:52:39] <Tempt> aaah
[09:52:48] <Tempt> frypan roasting is a risky proposition
[09:52:54] <Tempt> popper is much better.
[09:52:58] <e^ipi> evidently so
[09:53:05] <Tempt> Tried the Yrg yet?
[09:53:10] <WickedWicky> did you try the coffee?
[09:53:11] <WickedWicky> yea
[09:53:42] <e^ipi> no, i'm working on the crappy ones first so that I can ruin them without too much worry before I move on to the ones that I will actually care more about
[09:53:49] <Tempt> try pulling a shot right after roasting with no outgassing time
[09:53:58] <Tempt> and watch the crema spray out your portafilter
[09:54:15] <e^ipi> btw, taste-wise, why degass?
[09:54:16] *** Fish has joined #opensolaris
[09:54:35] <e^ipi> what sorts of flavours does it leave?
[09:54:52] <Fish> hello
[09:55:20] <Tempt> oh, the espresso won't pull properly without a day of rest.
[09:55:28] <Tempt> But you'll get some fun results immediately after roast.
[09:56:08] <WickedWicky> now I want a coffee.
[09:56:08] <WickedWicky> brb
[09:56:29] <e^ipi> Tempt: & in the cup ( french press, etc )?
[09:57:30] *** yongsun has quit IRC
[09:57:31] <e^ipi> I can notice a bunch of flavors, and i'm not sure what's due to variatals, my (lack of) roasting skill, or age
[10:00:03] *** dmarker has quit IRC
[10:00:17] *** dmarker has joined #opensolaris
[10:00:53] <Tempt> haven't tried home roasted with press or similar.
[10:01:03] <e^ipi> hmm
[10:01:04] <Tempt> Did one roast and decided it was too much effort.
[10:01:05] *** simford has quit IRC
[10:01:12] *** RainDoctor has joined #opensolaris
[10:01:29] <e^ipi> I can't find any notes anywhere as to why you should bother degassing if you're going to throw it in a press
[10:01:41] <e^ipi> it's just taken as gospel that you leave it sit for 24 hours
[10:01:51] <Tempt> like so many thinks in the world of coffee
[10:02:16] <Tempt> Anyway, I'm not expert on homeroasting, so I'd say experiment and find out what works for you and ignore the dogma.
[10:02:34] *** gavagai__ has joined #opensolaris
[10:02:49] <e^ipi> yeah, that was the plan
[10:02:52] <trochej> ookay
[10:02:57] <trochej> The server is up
[10:03:05] <trochej> STupid linux thingy
[10:03:15] <trochej> Or grub
[10:03:17] <trochej> Or whatever
[10:18:42] <WickedWicky> I'll have to reinstall my box too
[10:19:02] <WickedWicky> Sep 30 08:35:54 lilith tmpfs: [ID 518458 kern.warning] WARNING: /tmp: File system full, swap space limit exceeded
[10:19:03] <WickedWicky> Sep 30 08:36:03 lilith last message repeated 11125 times
[10:19:03] <WickedWicky> Sep 30 08:36:06 lilith syslogd: malloc failed: dropping message: Resource temporarily unavailable
[10:19:03] <WickedWicky> Sep 30 08:36:03 lilith tmpfs: [ID 518458 kern.warning] WARNING: /tmp: File system full, swap space limit exceeded
[10:19:03] <WickedWicky> Sep 30 08:40:30 lilith last message repeated 33356 times
[10:19:04] <WickedWicky> Sep 30 08:41:51 lilith tmpfs: [ID 518458 kern.warning] WARNING: /tmp: File system full, swap space limit exceeded
[10:19:07] <WickedWicky> Sep 30 08:45:05 lilith last message repeated 844 times
[10:19:09] <WickedWicky> this doesn't really look sexy
[10:19:14] <Tempt> haha
[10:19:17] *** RainDoctor has quit IRC
[10:19:18] <Tempt> out of memory?
[10:19:21] <WickedWicky> no
[10:19:21] <WickedWicky> hehehe
[10:19:25] <WickedWicky> there is enough memory
[10:19:28] <Tempt> or something gone sour filling /tmp?
[10:19:39] <WickedWicky> but you really shouldnt compile the onnv-gate with SS12 and think it'll run pretty
[10:19:41] * Tempt thinks WW will learn about constraining tmpfs sizing
[10:20:23] <WickedWicky> 2GB of RAM, 2GB of swap, you compile the onnv-gate with sun studio 12, you bfu, think "so that didnt go quite as bad as ppl say it would,they are on crack"
[10:20:26] <WickedWicky> then, lo and behold
[10:20:32] <WickedWicky> apache wont start and dies with SEGV
[10:20:37] <Tempt> haha
[10:20:46] <Tempt>  btw
[10:20:50] <Tempt> pebbles is down
[10:20:54] <WickedWicky> you think: ok, lets compile onnv-gate then with sun studio 11, see if it is an ON74 thing
[10:20:57] <WickedWicky> no
[10:20:58] <WickedWicky> lilith is down
[10:21:02] <WickedWicky> which is the database server
[10:21:04] <WickedWicky> :p
[10:21:05] <Tempt> aah
[10:21:05] <Tempt> mysql
[10:21:06] <Tempt> indeed
[10:21:15] <WickedWicky> pebbles is the most stable machine in this house :P
[10:21:59] <WickedWicky> but serious, I learnt a lot by volunteerly breaking stuff
[10:22:09] * WickedWicky is using truss/mdb / pstack now
[10:22:22] <WickedWicky> just to figure out why a compile would eat all my memory
[10:22:48] *** gavagai_ has quit IRC
[10:23:13] *** smtms has joined #opensolaris
[10:23:23] <WickedWicky> and thank god for ZFS which you can export, reinstall your system and then import again
[10:23:51] <Doc> yes, unlike every other filesystem which supports exactly the same thing
[10:24:00] <Doc> wait... that doesnt make sense
[10:24:13] <WickedWicky> hm?
[10:24:14] <WickedWicky> no no
[10:24:18] <WickedWicky> I used to have one big /
[10:24:23] <WickedWicky> that's what i mean :)
[10:24:35] <Doc> right, so this actually has nothing at all to do with ZFS
[10:24:46] <WickedWicky> I guess, still, I am happy I wont lose data
[10:24:49] <Doc> not saying it's not a good thing, just not overly unique :)
[10:24:52] <WickedWicky> leave me happy for a second or two
[10:24:56] *** estibi_ has joined #opensolaris
[10:26:01] *** Mdx4 has joined #opensolaris
[10:28:21] *** RainDoctor has joined #opensolaris
[10:31:21] *** gavagai_ has joined #opensolaris
[10:31:24] *** sahafeez has quit IRC
[10:36:53] *** estibi has quit IRC
[10:44:55] *** gavagai__ has quit IRC
[10:48:30] *** tombhad has joined #opensolaris
[10:49:17] *** gavagai__ has joined #opensolaris
[10:50:40] *** Shiv_1 has joined #opensolaris
[10:51:59] *** bengtf has quit IRC
[10:53:03] *** RainDoctor has quit IRC
[11:00:04] *** gavagai__ has quit IRC
[11:00:45] *** Shiv_1 has quit IRC
[11:06:27] *** gavagai_ has quit IRC
[11:10:24] *** andyshack has joined #opensolaris
[11:12:47] *** bengtf has joined #opensolaris
[11:13:02] <jteo> mmm. it seems that for all intent and purposes, IPS has been chosen.
[11:13:47] <jteo> oops. UPS.
[11:14:24] <sickness> ? :P
[11:15:14] <jteo> the packaging system.
[11:15:31] <WickedWicky> actually it's DHL here
[11:15:40] <WickedWicky> unless you dont mean those kind of packages
[11:20:30] <jteo> software.
[11:24:14] *** emiel_ has joined #opensolaris
[11:30:47] *** woei has joined #opensolaris
[11:33:46] *** Gekz has quit IRC
[11:34:28] *** Gekz has joined #opensolaris
[11:36:54] *** Gekz has quit IRC
[11:37:09] *** Gekz has joined #opensolaris
[11:37:28] *** Gekz has joined #opensolaris
[11:41:03] *** Gekz is now known as Gekkko
[11:42:08] *** Gekkko is now known as Gekz
[11:43:27] *** smesjz has joined #opensolaris
[11:43:47] <woei> I've found some blog postings about doing a pxe boot and then do a network install for x86 opensolaris
[11:44:03] <woei> however, they all asume you already have a solaris machine, with sun's dhcp server, running
[11:45:04] <woei> is there an easy way to do a network boot and install if you're serving a pxe boot image from, say, a bsd tftpd and dhcpd ?
[11:45:42] <sickness> yeah, it's perfectly doable, search on google, I've seen people do that from linux, freebsd, and even osx
[11:46:26] <woei> okay, mustn't be giving the correct search terms
[11:46:27] <sickness> you just need to configure dhcp and tftp in the right way, and nfs export the install directory, that's it, for better compatibility, isc dhcp and tftp-hpa are adviced
[11:46:44] <woei> okay, thanks for the confirmation
[11:47:17] <sickness> yw
[11:47:20] <WickedWicky> I think they mention the ISC dhcp server cause it is very easy to pass extra options to the DHCP client, like pboot path and all
[11:47:50] <WickedWicky> (and cause it's what comes with the OS)
[11:48:13] <WickedWicky> ok..
[11:48:18] <WickedWicky> let's see how this goes now
[11:48:19] <WickedWicky> brb
[11:51:23] <WickedWicky> yay, MySQl runs again
[11:51:42] *** Mdx4 has quit IRC
[11:55:48] <smesjz> Not directly related to OpenSolaris but i am downloading OpenSolaris DVDs using the Sun Download manager on win32. When i had restarted my computer, the download manager was unable to resume the download as it has status 'Stopped'. Is this normal behaviour or not?
[11:57:13] *** estibi_ has left #opensolaris
[11:59:09] <Doc> heh.. SDLC
[12:03:20] *** jteo has quit IRC
[12:05:55] *** NikolaVeber has joined #opensolaris
[12:17:21] *** blindfish has joined #opensolaris
[12:18:27] *** Gekz has quit IRC
[12:22:43] *** jte1 has joined #opensolaris
[12:22:57] *** jte1 is now known as jteo
[12:24:59] <JWheeler> The age old quetion. How do I kill a stale nfs mount? The server has been rebooted, I can mount the same path to another directory locally, but my old one is locked solid
[12:25:20] *** toblun has joined #opensolaris
[12:27:30] *** emiel_ has quit IRC
[12:28:35] *** Gekz has joined #opensolaris
[12:28:38] *** Pietro_S has joined #opensolaris
[12:58:19] *** woei has quit IRC
[13:01:12] *** Mdx4 has joined #opensolaris
[13:06:41] *** snuff-home has quit IRC
[13:11:44] *** Chihan has joined #OpenSolaris
[13:16:51] *** Chihan has quit IRC
[13:17:00] *** nostoi has quit IRC
[13:30:00] *** __hsilva has joined #opensolaris
[13:37:21] *** ichigo has joined #opensolaris
[13:42:34] *** _hsilva has quit IRC
[13:42:41] *** __hsilva is now known as _hsilva
[13:43:57] *** JWheeler has quit IRC
[13:47:39] *** kFuQ has joined #opensolaris
[13:50:30] *** Gekz is now known as Gekz[PDA]
[13:57:52] *** Mdx4 has quit IRC
[14:01:20] *** JWheeler has joined #opensolaris
[14:05:15] *** kFuQ has quit IRC
[14:11:45] *** Mdx4 has joined #opensolaris
[14:12:06] *** kFuQ has joined #opensolaris
[14:12:18] *** Fullmoon has joined #opensolaris
[14:13:16] *** bengtf has quit IRC
[14:16:02] *** master_of_master has joined #opensolaris
[14:16:11] *** jteo has quit IRC
[14:17:33] *** cypromis has joined #opensolaris
[14:17:42] *** Odin- has joined #opensolaris
[14:26:36] *** Syndelor has joined #OpenSolaris
[14:26:50] *** mcnamarabrian has joined #opensolaris
[14:28:19] *** mcnamarabrian has quit IRC
[14:28:32] *** master_o1_master has quit IRC
[14:32:49] *** bengtf has joined #opensolaris
[14:33:53] *** SeSe has quit IRC
[14:40:13] *** yarihm has joined #OpenSolaris
[14:46:28] *** stuart_ has joined #opensolaris
[14:47:42] <stuart_> can somebody help with update process?
[14:47:56] <trygvis> just ask
[14:49:05] *** stuart_ has quit IRC
[14:49:20] *** stuart_ has joined #opensolaris
[14:49:23] <stuart_> the update icon in the gnome panel says system not registered. When i select register nothing happens (other than the software to manage patches comes up)
[14:49:49] <quasi> which solaris version?
[14:50:10] <stuart_> latest developer build -- i think it's nv70
[14:50:21] <stuart_> in any case 9/07
[14:50:23] <quasi> afaik, there's no updates for it
[14:50:24] *** dpn` has quit IRC
[14:50:42] <quasi> you get those by grabbing the next nv
[14:51:05] <quasi> I think it is the same for sxde, but I'm not 110% sure
[14:51:15] <trygvis> yeah, I think that is correct
[14:51:21] <trygvis> there are no updates for the nv builds
[14:51:29] <stuart_> oh ok, so in effect no updates are available short of installing the next build?
[14:51:36] <trygvis> right
[14:51:39] <stuart_> nv=?
[14:52:02] <trygvis> nv70 is nv build 70
[14:52:07] <quasi> nv ~ SXCE which has gotten to 73 according to /topic
[14:52:20] <stuart_> is there a free version i can download that does the update thing
[14:52:44] <stuart_> actually i just downloaded for fun to see if it could replace my linux desktop
[14:53:48] <trygvis> then just stick to b70 for now, I doubt there is a whole lot of changes since 70
[14:54:02] <trygvis> but when you do want to upgrade, look at live upgrade
[14:54:45] <stuart_> im assuming live upgrade upgrades to a new build without me having to reinstall from scratch?
[14:55:17] <trygvis> sort of, yes
[14:55:23] <stuart_> is it "free"
[14:55:26] <trygvis> yes
[14:55:58] <stuart_> to find out about it i just search "live upgrade" on sun.com site?
[14:56:01] <trygvis> http://hell.jedicoder.net/?p=39
[14:56:07] <trygvis> yeah, or just google
[14:56:15] <stuart_> ok, thanks
[14:56:31] <trygvis> I think that is the page I used when I did my first LU
[14:56:34] <stuart_> so i just keep an eye out for new builds and then use the live upgrade
[14:56:57] <trygvis> or was it this one? http://www.sun.com/software/solaris/howtoguides/liveupgradehowto.jsp
[14:56:59] <trygvis> yep
[14:57:24] <Doc> hmm.. i think i helped write that
[14:57:28] <stuart_> k, thanks for your help
[14:57:29] <quasi> http://opensolaris.org/os/community/on/flag-days/71-75/ has a short summary of changes
[14:59:29] *** stuart_ has quit IRC
[15:00:04] *** andyshak has joined #opensolaris
[15:06:20] *** Pietro_S has quit IRC
[15:13:50] *** moo_cow has joined #opensolaris
[15:15:16] *** andyshack has quit IRC
[15:15:50] *** andyshak is now known as andyshack
[15:17:04] *** Mdx4 has quit IRC
[15:17:14] *** Fish has quit IRC
[15:19:32] *** Shiv_1 has joined #opensolaris
[15:23:20] *** medar has quit IRC
[15:26:40] *** medar has joined #opensolaris
[15:29:51] *** perlmonk has joined #opensolaris
[15:41:44] <Shiv_1> A pkg-discuss mailing list has been started !
[15:42:15] <Shiv_1> Atlast some discussions related to the packaging system are happening on a osol alias
[15:43:39] <Shiv_1> Anyone here tried the prototype?
[15:47:28] *** Gekz[PDA] is now known as gekz
[15:47:45] *** gekz is now known as Gekz[PDA]
[15:52:36] *** Odin- has quit IRC
[16:02:43] *** dunc has quit IRC
[16:06:46] *** dunc has joined #opensolaris
[16:08:56] *** Shiv_1 has quit IRC
[16:14:04] *** Gekz[PDA] is now known as Gekz[sleep]
[16:26:15] *** linma has joined #opensolaris
[16:30:02] *** linma has quit IRC
[16:33:37] *** insomnia has quit IRC
[16:40:10] *** logic has joined #opensolaris
[16:41:48] <tomww> Shiv_1: no, not tried.
[16:54:42] *** logic_ has quit IRC
[16:54:53] *** the_reuper has joined #opensolaris
[16:57:50] <quasi> hrm, how do I convince sun cluster to let go of my disks and let me deport the disk group
[16:58:53] <quasi> ah, read the docs ;)
[16:59:26] <Doc> you should never deport disk groups in cluster
[16:59:37] <Doc> SC should do it for you
[17:00:00] <Tempt> Doc's right.
[17:00:18] <quasi> http://docs.sun.com/app/docs/doc/819-0580/6n30eahbr?a=view says deport
[17:00:39] <quasi> I want to move the diskgroup out of this cluster and into another
[17:00:50] *** mineye has joined #opensolaris
[17:01:13] <Tempt> aaah.
[17:01:23] *** ashner has quit IRC
[17:01:37] <Doc> tempt: you mean i'm right, or the docs on docs.sun.com are right? :)
[17:01:55] <Doc> quasi: so remove it from the cluster and add it on the other
[17:01:56] <Tempt> Well, I didn't realise he was trying to remove his diskgroup from the cluster.
[17:02:01] <Tempt> That doesn't happen too often.
[17:02:01] <Doc> it will deport/import it for you in the process
[17:02:22] <Doc> take it offline, the use scsetup or whatever it's called to remove it
[17:02:32] <quasi> oh wait, I'm reading the doc wrong
[17:02:45] * Doc is often the wrong doc
[17:02:50] *** the_reuper has left #opensolaris
[17:02:59] <WickedWicky> lol
[17:03:10] <quasi> ;)
[17:04:07] <quasi> they say import / deport if you want to use the diskset outside the cluster
[17:04:14] <quasi> doh!
[17:05:03] <Tempt> got the right documentation now?
[17:05:46] <Tempt> http://docs.sun.com/app/docs/doc/819-0580/6n30eahbf?a=view
[17:10:12] <quasi> scswitch: service is busy
[17:11:27] <WickedWicky> mounted still?
[17:11:42] <quasi> yes
[17:11:57] <WickedWicky> so you'll have to offline the resource first I think
[17:12:06] <WickedWicky> oh wait
[17:12:09] <WickedWicky> that's what you're doing
[17:12:11] <WickedWicky> nevermind
[17:14:15] *** Fish has joined #opensolaris
[17:18:29] <quasi> Tempt: got any ideas?
[17:18:51] <Tempt> so all the resources are down
[17:19:07] <Tempt> and there aren't any hastorageplus objects left
[17:19:08] <Tempt> etc
[17:19:13] <Tempt> it's all clean except the dg?
[17:19:55] <quasi> resources and resource groups are offline
[17:21:22] <quasi>   Device group status:        roamdg              Online
[17:21:23] <quasi>   Device group status:        oradg               Online
[17:21:28] <quasi> is all I have online
[17:21:35] <Tempt> and you can't offline them?
[17:21:40] <quasi> no
[17:21:48] *** perlmonk has quit IRC
[17:21:52] <Tempt> check mounted filesystems on both nodes?
[17:22:08] <quasi> all are mounted
[17:22:13] *** perlmonk has joined #opensolaris
[17:22:40] <Tempt> umount them then.
[17:22:45] *** dunc_ has joined #opensolaris
[17:23:16] <quasi> ehrm, I was going to say unmounted, but I actually missed one
[17:23:47] <quasi> much better now, thanks
[17:24:06] *** dunc has quit IRC
[17:32:39] *** pinskian has joined #opensolaris
[17:32:55] *** pinskian has left #opensolaris
[17:34:37] *** trasz has joined #opensolaris
[17:34:49] <trasz> this question is _not_ meant as a flame: why is cde on suns seem so slow?  no matter where, sparcstation 2, ultra 5 or blade 1000, the mouse lags a little and the rest of the gui does that too.
[17:34:59] <trasz> i thought it was a cde problem, but it feels totally snappy on 132mhz pa-risc.
[17:35:03] *** gavagai__ has joined #opensolaris
[17:40:24] *** m0le has joined #opensolaris
[17:41:11] *** Syndelor has quit IRC
[17:41:51] *** vmlemon has joined #opensolaris
[17:42:33] <smtms> trasz, I think it is scheduler related
[17:43:46] <trasz> smtms: any way to tune it?
[17:44:07] <trasz> smtms: sun was a primarily workstation company, iirc.
[17:44:27] <trasz> smtms: i'd expect scheduler to 'do the right thing' for workstations by default.
[17:45:19] <Tempt> smtms: Why do you say scheduler?
[17:45:24] *** tombhad has quit IRC
[17:45:54] <smtms> trasz, because I use Solaris 10 and the interactivity is not good and I've read somewhere of a complicated way to change the scheduler
[17:45:57] *** tombhad has joined #opensolaris
[17:46:16] *** JoergB has quit IRC
[17:46:31] <trasz> smtms: i seem to remember that several years ago i even tried giving Xsun realtime priority.  mouse still lagged.
[17:47:02] <Tempt> Xsun will actually bump the priority of the focussed window, to improve application response
[17:47:17] <trasz> smtms: another thing is, mouse lagged in exactly the same way under sunos 4, bsd-based.
[17:47:18] <Tempt> if you're on an old SPARCstation and the mouse is lagger, that's the mouse architecture at work.
[17:47:27] <e^ipi> is there a way to programaticaally turn on the micro timer?
[17:47:42] <e^ipi> ie, so one doesn't have to run ptime
[17:47:43] <smtms> the BSD scheduler is not interactive
[17:48:16] <trasz> Tempt: you mean, the mouse hardware?  i seem to remember some hack made to sun optical mouse to raise resolution, or something.
[17:48:57] <Tempt> I think the data rate of the mouse was very low.
[17:49:13] <Tempt> the hack, IIRC, was to increase the baud rate.
[17:50:02] <smtms> trasz, how does it fare under load? does the mouse start skipping?
[17:54:20] <trasz> smtms: you mean, under solaris?
[17:54:53] <smtms> trasz, yes
[17:54:55] <trasz> smtms: don't remember right now, but under some _really_ big load it probably will.  i was asking about performance with zero load, however.
[17:57:52] *** Syndelor has joined #OpenSolaris
[17:58:55] <e^ipi> oh, microstate accounting is on by default
[17:58:57] <e^ipi> that's cool
[17:59:17] <Tempt> of course it is.
[17:59:44] <Tempt> even loadavg is calculated using microstate
[18:00:41] *** rmorse has joined #opensolaris
[18:03:27] *** mineye has quit IRC
[18:04:56] <e^ipi> Tempt: it's not turned on in 9 by default
[18:05:22] <e^ipi> & i'm writing something that needs it
[18:09:13] *** sioraiocht_ has joined #opensolaris
[18:10:32] *** sioraiocht has quit IRC
[18:11:07] <e^ipi> does 9 have a way of turning on other than hacking it with ptitme ?
[18:24:33] *** ciph3r has joined #OpenSolaris
[18:26:52] *** dunc_ has quit IRC
[18:29:34] *** estibi has joined #opensolaris
[18:35:06] *** bunker has joined #opensolaris
[18:40:44] *** agrebnev has joined #opensolaris
[18:41:14] <agrebnev> Hi All
[18:45:49] *** gavagai_ has joined #opensolaris
[18:47:41] *** rmorse has quit IRC
[18:49:57] *** trendoid has joined #opensolaris
[18:50:32] *** Netwolf_ is now known as Netwolf
[18:52:23] *** trendoid has quit IRC
[18:52:59] *** bengtf has quit IRC
[18:54:23] <jamesd> hi
[18:55:23] *** trendoid has joined #opensolaris
[18:56:12] *** locy has joined #opensolaris
[18:56:56] <locy> nabend zusammen
[19:00:55] *** kadath__ has quit IRC
[19:03:38] *** gavagai__ has quit IRC
[19:04:08] *** agrebnev_ has joined #opensolaris
[19:04:33] *** agrebnev_ has quit IRC
[19:04:50] *** agrebnev_ has joined #opensolaris
[19:06:05] *** Shiv_1 has joined #opensolaris
[19:08:12] *** agrebnev_ has quit IRC
[19:11:50] *** andyshack has quit IRC
[19:15:05] *** jollyd has joined #opensolaris
[19:15:08] *** bengtf has joined #opensolaris
[19:15:18] <jollyd> hi
[19:16:15] *** cypromis has quit IRC
[19:16:21] <locy> jollyd du arbeitest auch mit opensolaris?
[19:16:23] *** sommerfeld has joined #opensolaris
[19:16:24] *** ChanServ sets mode: +o sommerfeld
[19:16:43] <gavagai_> hey
[19:17:49] *** deather_ has joined #opensolaris
[19:18:15] 
[19:18:22] <jollyd> locy: yes i do. on my server and an old Ultra2
[19:18:47] <jollyd> but i am a beginner on solaris
[19:19:03] <locy> ok me too
[19:19:11] <jbk> locy: any error messages?
[19:19:19] <jollyd> locy: i nerver encountered such a thing
[19:19:24] *** agrebnev has quit IRC
[19:19:42] <locy> no only he didnt find a boot file
[19:21:34] <locy> but on my main pc he boot! my second pc is a ibm thinkcentre
[19:24:46] *** agrebnev has joined #opensolaris
[19:25:18] *** blindfish has quit IRC
[19:27:41] *** charlesnw has joined #opensolaris
[19:29:41] *** agrebnev has left #opensolaris
[19:30:06] *** agrebnev has joined #opensolaris
[19:30:09] *** TheZeroX has joined #opensolaris
[19:30:27] <agrebnev> Hi All
[19:33:24] <agrebnev> Are anybody here?
[19:34:24] *** deather has quit IRC
[19:34:25] *** blindfish has joined #opensolaris
[19:34:41] <smtms> agrebnev, privet
[19:34:57] <smtms> agrebnev, it should be obvious that there are some people here
[19:35:22] *** Shiv_1 has quit IRC
[19:35:28] <agrebnev> Sorry I use this IRC client for the first time.
[19:35:48] <agrebnev> And this channel too
[19:36:03] <agrebnev> I thought it is more active
[19:36:44] *** suhail has joined #opensolaris
[19:37:03] <suhail> hi
[19:37:08] *** trasz has quit IRC
[19:37:30] <agrebnev> I have some questions about Solaris. Who have time to answer me?
[19:37:49] <asyd> just ask them (and wait)
[19:37:57] *** ChanServ sets mode: +o asyd
[19:42:07] *** suhail has quit IRC
[19:45:24] <agrebnev> I have installed my Solaris XDE 5/07 for the first time into my VMWare Player. I have problems with terminal. E.g. my arrow buttons and Delete button do not work. Please give to me advice.
[19:46:47] *** trasz has joined #opensolaris
[19:49:53] <smtms> agrebnev, are you logged in as root on the console?
[19:50:25] <agrebnev> yes
[19:50:44] <agrebnev> TERM=xterm
[19:53:37] <jamesd> run bash
[19:56:59] <agrebnev> Thanks. It works. What is the shell which I use by default?
[19:58:51] <trendoid> Why does devfsadm not create the disk symlink for me, i.e. c0t0d0. All I get is the pX and sX symlinks. I'd like to use the complete disk as a ZFS mirror.
[19:58:57] *** derchris^ has quit IRC
[19:59:06] *** derchris^ has joined #opensolaris
[20:00:07] <kjetilho> trendoid: so use s2?
[20:00:20] <kjetilho> in SPARC, there is no device for the whole disk
[20:00:59] <trendoid> I'm on x86. Yes s2 is the whole disk. Not used to that. Will it work?
[20:01:14] *** charlesnw has quit IRC
[20:01:47] <kjetilho> trendoid: that's the customary way
[20:02:00] <jbk> if you omit the slice w/ zfs, it will use the whole disk
[20:02:25] <trendoid> jbk: But a symlink has to exist first.
[20:02:30] <jbk> no
[20:02:37] <trendoid> I'll try now.
[20:02:41] <jbk> it will assume s2
[20:03:29] <WickedWicky> erm
[20:03:49] <WickedWicky> if you want to use the entire disk with zfs you should use cxdy
[20:04:01] <WickedWicky> like, c1d0
[20:04:16] <WickedWicky> example:
[20:04:17] <WickedWicky>        NAME        STATE     READ WRITE CKSUM
[20:04:17] <WickedWicky>         lilith_raid  ONLINE       0     0     0
[20:04:17] <WickedWicky>           mirror    ONLINE       0     0     0
[20:04:17] <WickedWicky>             c4d0    ONLINE       0     0     0
[20:04:17] <WickedWicky>             c3d0    ONLINE       0     0     0
[20:04:22] <WickedWicky> oh
[20:04:27] <WickedWicky> jbk already mentioned that
[20:04:27] <WickedWicky> lalaa
[20:04:32] * WickedWicky is slow today
[20:04:34] <trendoid> jbk:  You are correct. :)
[20:05:25] *** sahafeez has joined #opensolaris
[20:05:33] <trendoid> WickedWicky: Right you are. The example you give is for an IDE disk.
[20:05:47] <WickedWicky> sata, in this case
[20:06:12] <jamesd> WickedWicky, can i paste my zfs pool?  it "only" has 15 drives in it....
[20:06:24] <WickedWicky> jamesd, knock yourself out
[20:06:31] <WickedWicky> :P
[20:06:54] *** charlesnw has joined #opensolaris
[20:07:08] <WickedWicky> oh christ
[20:07:09] <trendoid> I'm still getting used to devfs. I'm used to more static situations. :)
[20:07:10] <WickedWicky> he does it
[20:07:11] <WickedWicky> lol
[20:07:16] <trendoid> I'm happy.
[20:07:21] <trendoid> Thanks.
[20:07:43] <WickedWicky> jesus jamesd, how much terra is that?
[20:08:07] <jamesd> WickedWicky, 148 GB, those are  18Gb and 9GB  drives..
[20:08:33] <WickedWicky> cool
[20:09:30] <agrebnev> How can I configure loader in order to have facility to load Solaris with and without XWindows?
[20:10:15] <g4lt-sb100> technically, it isn't solaris without a GUI ;P
[20:10:31] <jollyd> i wonder if somebody can give me a hint.     i try to compile paraview w/ the gcc compiler packaged by blastwave (cmake does not let me use suncc).      the sun linker is used.      and i get some "relocation error:   symbol has been discarded with discarded section: .gnu.linkonce.t".. but i'm not a programmer and i already have this problem w/ other sources... where can i find some hint to solve this kind of problem ?
[20:11:56] *** sch has joined #opensolaris
[20:12:49] <axisys> how do I do a backtrace of a coredump using mdb?
[20:13:00] <axisys> i took a core dump using gcore
[20:13:18] <axisys> now I need to find out where is the process hanging
[20:13:47] <sch> pstack works on core files (as well as live)
[20:14:30] <sch> but $C will show the stack trace in mdb
[20:14:50] <sch> (as will ::stack)
[20:15:42] <agrebnev> Are there some package manager in Solaris? I want to install mc.
[20:16:46] <sch> Not as an integrated; blastwave.org offers pkg-get and a collection of software.
[20:18:17] <sch> blastwave has mc-4.6.1
[20:18:34] *** deather_ is now known as deather
[20:19:43] <axisys> i have 17 threads.. http://opensolaris.pastebin.ca/720785
[20:19:55] <axisys> is there a way to tell which one is causing the hang?
[20:20:58] <axisys> mdb core ::stack shows only shows this.. can;t tell which thread it belongs to http://opensolaris.pastebin.ca/720787
[20:23:08] <agrebnev> sch, thanks
[20:27:40] <sch> It's the main thread.  Use ::walk thread | ::findstack to see the complete list of thread stacks.
[20:27:42] <agrebnev> Does somebody begin to make package of http://open-vm-tools.sourceforge.net/ ? Who knows?
[20:29:20] <axisys> sch: cool!! thanks a lot
[20:42:29] <agrebnev> How to enable the Home and End buttons in bash too?
[20:46:46] <madhatter> agrebnev: I think it's more a feature of xterm but of the shell, but I am not sure right now.
[20:47:24] <kjetilho> no, it's the shell's responsibility to recognise the escape sequence emitted
[20:47:47] <madhatter> Okay, so I might be wrong ;)
[20:48:06] <madhatter> I thought I did stuff like that in the ~/.Xdefaults
[20:48:08] <kjetilho> although XTerm can send two different types of escapes for Home and End
[20:48:29] <kjetilho> ^[[H (vt102 mode, default) ^[[1~ (vt220 mode)
[20:48:38] <madhatter> But I am no fan of Home and End, as there is ctrl+a and e
[20:49:03] <kjetilho> me too :)  I have a Happy Hacking Keyboard, so the keys don't even exist
[20:50:02] <madhatter> I am still searching for a good reseller for those. Must fit almost the size of my 12" iBook's keyboard. I think I can press some keys at the same time to get Home and End
[20:50:11] <WickedWicky> what on earth is a happy hacking keyboard
[20:51:05] <agrebnev> kjetilho, do you offer to me to change my env var TERM to vt220 ?
[20:51:06] <madhatter> A very small keyboard. With just the keys you *really* need
[20:51:57] <madhatter> WickedWicky: http://www.xemacs.org/~mrb/nhk/happy-hacking.jpg
[20:52:36] <WickedWicky> haha
[20:52:37] <WickedWicky> cool
[20:52:45] <WickedWicky> where do I buy this?
[20:53:13] <kjetilho> agrebnev: I don't think it will help.  but also try Contrl+Left button
[20:53:25] <madhatter> I know only of expensive sources here in Germany. You are from the Netherlands?
[20:53:33] <WickedWicky> ja
[20:54:03] <WickedWicky> I will be in Saarbrucken in some weeks
[20:54:09] <kjetilho> I really should get the version without the text on the keys :)
[20:54:29] <madhatter> kjetilho: You should :)
[20:54:44] <madhatter> There is a light version which has cursor keys, too.
[20:55:20] <madhatter> WickedWicky: Hm, I only tried to order one online, but iirc they wanted 150 Euros for it
[20:55:27] <quasi> kjetilho: nah, the black keyboard is nice - no risk of people stealing your desk at work ;)
[20:55:56] <madhatter> WickedWicky: And I have never been to Saarbrucken.
[20:55:59] <quasi> order one in .us - with the current $ rate, it can't be all that expensive ;)
[20:56:35] <WickedWicky> if it goes on like this the canucks will break even with the US$
[20:57:00] <madhatter> quasi: True, but you might have to add custom payments if it comes bad.
[20:57:15] <quasi> http://pfuca-store.stores.yahoo.net/haphackeylit1.html - $69
[20:57:23] <madhatter> No idea if that might happen in .nl, too.
[20:57:25] <kjetilho> WickedWicky: CAD passed USD last week
[20:57:34] <WickedWicky> really? wow
[20:58:09] <quasi> yeah, it did
[20:58:10] <madhatter> quasi: That's cheap
[20:58:30] <quasi> madhatter: yeah - compared to the pro with blank keys at $259
[20:58:51] <kjetilho> the Happy Hacking's feel is modeeled after the Type 3 keyboard, if any of you remember that one
[20:59:41] *** Triskelios has joined #opensolaris
[20:59:43] <madhatter> quasi: Wow, that just can be because of the 'pro' in the name.
[21:00:43] <madhatter> kjetilho: No, I don't. From when is that?
[21:00:54] <agrebnev> Are there something like
[21:00:55] <agrebnev> cat /proc/cpu_info
[21:00:57] <agrebnev> cat /proc/mem_info
[21:00:58] <kjetilho> Sun3 (and possibly Sun2)
[21:00:59] <agrebnev> ?
[21:01:15] <trygvis> agrebnev: psrinfo
[21:01:19] <trygvis> and prtconf
[21:01:19] <kjetilho> Type 4 was the first with the connector which lasted until USB
[21:01:28] <madhatter> I am using a heavy, old IBM keyboard. Good to hang up pictures in the appartment. ;)
[21:01:54] <quasi> madhatter: and I'm sure it costs a bundle to remove the print
[21:02:43] <madhatter> Ya, sute
[21:02:46] <madhatter> *sure
[21:04:43] <madhatter> kjetilho: Just looked up pictures. I first typed on Type 4 keyboards on Suns
[21:05:01] <kjetilho> the Type 4 is very good, too.
[21:05:18] <kjetilho> I don't like the newer keyboards with their short travel length
[21:05:28] <kjetilho> all these laptop style keyboards drive me crazy
[21:06:07] *** moo_cow has quit IRC
[21:06:27] *** LeftWing has quit IRC
[21:07:00] *** LeftWing has joined #OpenSolaris
[21:08:17] <madhatter> You get used to it, but I really like the loud typing on my ibm keyboard. It's just not suitable for shared office space with co-workers. ;)
[21:11:20] *** sch has quit IRC
[21:17:11] <WickedWicky> clickety click
[21:25:38] *** paulf has joined #opensolaris
[21:25:55] *** gavagai_ has quit IRC
[21:28:06] <madhatter> Haha, ja
[21:30:24] <paulf> Hello
[21:31:22] *** Mdx4 has joined #opensolaris
[21:33:58] *** LeftWing has quit IRC
[21:34:01] *** LeftWing has joined #OpenSolaris
[21:36:23] <trochej> Hmmm
[21:36:39] *** charlesnw has quit IRC
[21:36:48] *** _dreams_ has joined #OpenSolaris
[21:38:38] * quasi always keeps a few spare ibm keyboards just in case the cow-orkers don't understand the value of keeping sounds off on their machines
[21:39:39] <quasi> I used to cut their speaker cables, but it is too hard nowadays when everybody has laptops
[21:40:48] <kjetilho> a cup of coffee into their speakers works wonders
[21:43:49] <trochej> kjetilho: ROTFL
[21:44:45] *** agrebnev has quit IRC
[21:45:04] <quasi> kjetilho: except that we have stinkpads - nowadays they have drain holes
[21:45:32] <madhatter> Do I need exactly b68 to use Solaris Cluster Express 7/07?
[21:45:49] <quasi> madhatter: I've been wondering about the same
[21:46:11] <quasi> madhatter: you should send a mail to the list asking (and save me the trouble ;)
[21:46:48] <madhatter> quasi: But you did not give it a try either?
[21:47:34] <quasi> madhatter: no, not yet - my x86 cluster test box has been stuck in the serverroom for about a week being used as a spare jumpstart box
[21:48:34] <madhatter> quasi: Maybe you can give me hint on another question then? What's the thing about a single-node cluster anyways?
[21:51:08] *** joideojed has joined #opensolaris
[21:51:21] *** WickedWicky has quit IRC
[21:51:25] *** joideojed is now known as WickedWicky
[21:52:38] <quasi> madhatter: just fun for testing
[21:53:21] <quasi> madhatter: when you're building cluster data services, it is nice to have somewhere that you can test them without needing shared storage or 3 machines
[21:53:49] <madhatter> quasi: In the howto for the installation of the single-node cluster is no special version of sxce mentioned.
[21:54:20] <quasi> madhatter: no, you could do single node on cluster 3.2 afaik
[21:54:47] <quasi> madhatter: but cluster express supports 32bit intel which regular cluster doesn't
[21:54:48] <madhatter> quasi: No, I ment about the b68 question
[21:55:22] <quasi> madhatter: no, but it is mentioned in the release notes for cluster express somewhere
[21:55:53] <madhatter> quasi: Ya, but I thought if it would be necessary it would be mentioned there either...
[21:56:32] <quasi> madhatter: ehrm, whatever - my brain is not capable of understanding - I've been at work for too long
[21:57:53] <madhatter> quasi: Hehe, okay
[21:59:08] *** jollyd has left #opensolaris
[21:59:40] * quasi has been fighting a cluster in production for hours
[21:59:47] * quasi lost
[22:00:56] <madhatter> quasi: I am sorry to hear that. What issues did the cluster have then?
[22:00:59] <WickedWicky> you didnt
[22:01:01] <WickedWicky> you asked for a timeout
[22:01:04] <WickedWicky> dont give up
[22:01:57] *** TheZeroX has quit IRC
[22:02:00] *** LeftWing has quit IRC
[22:02:03] *** LeftWing has joined #OpenSolaris
[22:04:17] *** peemus-home has joined #opensolaris
[22:08:03] *** Fish has quit IRC
[22:10:26] *** trendoid has quit IRC
[22:10:40] *** blindfish has quit IRC
[22:11:36] *** jafari has joined #opensolaris
[22:12:24] * quasi finishes the fallback and now turns to the task of finding food and a way out of this hellhole ;)
[22:12:39] <WickedWicky> sounds like the perfect project to end the day with
[22:12:44] <WickedWicky> go for nachos
[22:13:19] <quasi> work is literally in the middle of nowhere - I've got an hour on the train back
[22:13:29] <WickedWicky> jesus
[22:14:48] <quasi> I'll let $boss pay for a taxi
[22:15:11] * delewis has to make an hour commute to work for his new job
[22:15:12] <WickedWicky> least he can do
[22:15:22] <delewis> fortunately, my shift starts at 11PM, so no traffic. :-)
[22:15:30] <WickedWicky> that's good
[22:15:51] <WickedWicky> I travel an hour by public transport to work and an hour back, and I try to avoid rushhour
[22:15:52] *** kadath has joined #opensolaris
[22:16:01] * WickedWicky hates standing in the bus
[22:16:03] <delewis> yep.
[22:16:16] <delewis> rush hour in Memphis is awful, but I also get off 7AM, so I'm just barely missing it. :-)
[22:17:51] <delewis> then I've got class at 8AM for the next two months until I graduate
[22:21:37] *** locy has quit IRC
[22:21:40] <kjetilho> couldn't you find a McDonald's a bit closer to home?
[22:22:10] *** loky has joined #opensolaris
[22:22:19] <flyingparchment> can solaris terminate a vpn/ipsec connection?
[22:22:26] <flyingparchment> (gre+ipsec or so)
[22:22:26] <delewis> kjetilho: I'm sure I could have. But FedEx's World Technical datacenter isn't here, though. :-)
[22:23:04] <kjetilho> sounds nice
[22:24:41] *** Bartman007 is now known as The
[22:24:47] *** The is now known as The_Peasants
[22:25:43] *** kimc has left #opensolaris
[22:29:53] *** smesjz has quit IRC
[22:31:02] *** The_Peasants is now known as Bartman007
[22:31:21] *** movement has quit IRC
[22:31:28] *** movement has joined #opensolaris
[22:32:04] *** LeftWing has quit IRC
[22:32:17] *** LeftWing has joined #OpenSolaris
[22:38:56] *** Syndelor has quit IRC
[22:40:28] <sommerfeld> flyingparchment: solaris does ipsec but not gre.
[22:40:49] *** yarihm has quit IRC
[22:40:50] <flyingparchment> sommerfeld: is there anything similar that is supported?
[22:41:48] *** estibi has quit IRC
[22:41:52] <sommerfeld> ip over ipsec over ip tunnels work
[22:46:04] *** vmlemon has quit IRC
[22:49:07] *** cypromis has joined #opensolaris
[22:58:03] *** Gman has joined #opensolaris
[22:58:26] *** bunker has quit IRC
[23:00:53] *** nostoi has joined #opensolaris
[23:01:08] *** Ober has joined #opensolaris
[23:01:34] <Ober> snv_73 considered Solaris? or OpenSolaris? just want to make sure it's ok to refer to it as "opensolaris"
[23:02:14] <trygvis> it's opensolaris
[23:02:17] <jmcp> Ober: OpenSolaris - technically - is just the source code. Solaris is a released product from Sun, and
[23:02:58] <jmcp> Ober: http://whacked.net/2005/06/21/confused-so-was-i/
[23:03:04] <jmcp> explains the differences quite well
[23:03:15] *** kaiwai has joined #opensolaris
[23:05:27] *** estibi has joined #opensolaris
[23:12:56] <Ober> thanks. doing benchmarks. wanted to make sure I could call it opensolaris
[23:13:42] <elektronkind> it's more like "Solaris Express, Community Edition" aka SXCE
[23:14:02] <Ober> ok
[23:14:23] <elektronkind> "OpenSolaris" refers to the project itself, the community, and the source code
[23:14:43] <sickness> any news on getting smartctl to work with ata/sata disks under sxce? =)
[23:14:55] <elektronkind> think of SXCE (and SXDE - Developer Edition) as distributions from the OpenSolaris project
[23:15:04] <jmcp> sickness: no news on that at present
[23:15:07] <elektronkind> like Nexenta and Bellinix are as well
[23:15:17] <Ober> do they differ in the kernel ?
[23:15:34] <Ober> just don't want the usual "why didn't you benchmark X version of Y?"
[23:15:57] <sickness> jmcp: tnx
[23:16:25] <elektronkind> Ober: SXCE would be fine
[23:16:53] <Triskelios> Ober: just be aware that Solaris 10 is not based on the same codebase
[23:16:54] <elektronkind> but doing a SXDE might be "better" in some peoples' eyes
[23:17:08] *** loky has quit IRC
[23:17:36] <Ober> excellent thanks
[23:18:15] <Ober> I should expect zfs filesystem performance to be on a much better level thant he default ufs? as right now I am benchmarking sysbench/mysql on the default ufs / it created
[23:18:23] <trygvis>  what does ttymon do? and why would it use 100% CPU on my box?
[23:18:26] <delewis> depends on the workload.
[23:18:42] <flyingparchment> Ober: for innodb, zfs is still a fair bit slower than ufs
[23:18:53] <delewis> for database workloads, ZFS isn't quite up to par compared to other filesystems. The vdev cache enchancements that have already gone into Nevada should improve that, though.
[23:19:02] <delewis> the vdev cache enhancements didn't make it back to S10u4, though :-(
[23:19:03] <Ober> ok
[23:19:29] <Ober> just know I will hear "why didn't you try zfs!"
[23:19:30] <delewis> http://blogs.sun.com/erickustarz/entry/vdev_cache_improvements_to_help
[23:19:38] <sickness> jmcp: so, I must assume, no way to see the temperature of a pata/sata disk under snv72 (i86pc), right?
[23:19:45] <jmcp> sickness: correct
[23:19:48] <sickness> tnx
[23:21:25] <delewis> I'd still be hesitant about using ZFS in production for database workloads until the vdev_cache enhancements make it back to Solaris 10; however, in a testing environment, they're ideal, because of snapshots.
[23:21:58] <delewis> s/they're/its/
[23:22:56] <flyingparchment> Ober: you might want to look at vtune.sh, although i'm not sure whether it works on u4 (or whether it's still needed)
[23:23:10] <delewis> with the exception of checksumming most of what ZFS does is overhead for a transactional database, which already does COW in a sense.
[23:23:20] <Ober> ok will do thanks found the docs for solaris perf tuning for mysql
[23:25:31] <jbk> it would be interesting if there was direct access to the dmu interfaces -- instead of having two transaction interfaces (one in the database, one in the filesystem) let zfs take care of it
[23:25:34] <paulf> is ZFS pretty mature, or is it still evolving pretty quickly?
[23:25:48] <jbk> and the database could merely be an object set of zfs objects (tables and indexes)
[23:25:48] *** Drone has quit IRC
[23:25:53] <delewis> paulf: it's mature *and* evolving quickly.
[23:26:16] <delewis> mostly because ZFS's metadata structures were planned properly from the outset.
[23:26:32] <delewis> it's trivial to add new compression alogorithms and encryption because of that.
[23:26:55] <delewis> jks: interesting.
[23:27:00] <delewis> jbk*
[23:27:12] <jks> I'm always interesting
[23:27:38] <jbk> though even outside of that, having some sort of access to the transaction semantics even for zpl related operations would be neat i.e. begin_transaction; read, write, unlink, ....; commit;
[23:27:59] <delewis> jbk: yep, COW is huge overhead for something like Oracle, which already has rollback.
[23:28:12] <delewis> so not only do you have COW for the pending transaction, but the pending transaction in the rollback.
[23:28:49] <jbk> right now my understanding is each posix operation is its own transaction (which are grouped together, but not the same thing)
[23:30:01] <jbk> i believe there has been talk of eventually doing some (if not all) of those things, but still on the drawing board at this point..
[23:30:14] <paulf> I manage a netapp at work, and I must admit that the capabilities that ZFS bring are very apealing
[23:31:07] <kaiwai> hmm, cow
[23:31:09] <kaiwai> mooo
[23:31:12] <jbk> paulf: and cheaper :)
[23:31:24] <paulf> definitely :)
[23:32:47] <paulf> it could definitely bring very nice capabilities to smaller busineses
[23:33:18] <sommerfeld> delewis: so, one thing about zfs's COW is that they turn "random" writes into sequential writes.
[23:35:28] *** ACfromTX has joined #opensolaris
[23:37:07] <jbk> does WAFL still have limitations on the # of snapshots allowed?
[23:37:08] <sommerfeld> so if you set the recordsize or volblocksize to match the blocksize used by the database, zfs won't have to read the old zfs block contents to write
[23:38:34] <paulf> jbk:  I'm not certain...we don't normally keep that many snapshots
[23:42:09] <pjd-> jbk: Yes, I think it's 32, because they use reference counting for blocks using 32bit integer.
[23:42:25] <pjd-> (every bit represents one snapshot)
[23:43:45] <sommerfeld> zfs tracks snapshots using the block birth time (transaction group number).
[23:44:21] <sommerfeld> when zfs would free a block that's older than the newest snapshot, it doesn't actually free it.
[23:44:27] <pjd-> Yeah, zfs is smarter than WAFL here.
[23:44:57] <jbk> well that's what i was getting towards, i just don't know enough enough about WAFL to know if my info is still current
[23:45:04] <jbk> (apparently it is in that area at least)
[23:45:11] <sommerfeld> (the hairy part is deferred to snapshot deletion)
[23:46:33] *** cypromis has quit IRC
[23:52:12] *** nostoi has quit IRC
[23:55:29] *** mega has joined #opensolaris
[23:57:15] *** karrotx has quit IRC





top