July 12, 2011 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
July 12, 2011 [00:00:46] <DanaG> hmm, trying to change files via samba;[00:01:02] <DanaG> You require permission from MICROSERVER\dana to change these files.[00:02:12] *** DeanoC has quit IRC[00:03:38] *** DeanoC has joined #openindiana[00:03:39] *** ChanServ sets mode: +o DeanoC[00:03:47] <gnomad> DanaG: I generally maintain a separate install of Perl oh which I populate CPAN[00:03:55] <sergefonville1> if processes do not lose their contract, then why is it that when all processes are started in a schript and exit 0, SMF thinks they all exited[00:04:03] <gnomad> as a general rule, I leave the system Perl untouched.[00:07:42] <tsoome> how did you define the service?[00:08:39] <DanaG> I wouldn't know how to do that.[00:08:41] <sergefonville1> when I tell uwsgi to daemonize, it does not work (all processes exited) when I tell it, not to daemonize, the service and processes are fine[00:09:44] <sergefonville1> th only difference is that when they daemonize their parent becomes 1[00:09:54] <sergefonville1> that I know of[00:10:06] *** mikw has quit IRC[00:12:02] <richlowe> If you tell it to daemonize it forks setsids and exits[00:12:05] <richlowe> so it... exited.[00:12:21] <tsoome> check if you have <property_group name='startd' type='framework'> in the xml and what propval its setting there if you have[00:12:33] <richlowe> in general, if you're running it under SMF don't bother letting it daemonize[00:12:52] <richlowe> or tweak the service, as I assume tsoome is about to take you through.[00:15:11] *** gea__ has joined #openindiana[00:15:32] <tsoome> well, if you have samba installed, you can compare with its xml, as smbd is started with -D (demonize)[00:15:58] <tsoome> or some other service with daemon[00:17:04] <sergefonville1> <propval name='ignore_error' type='astring' value='core,signal' />[00:17:44] <tsoome> the difference are scripts which dont leave anything running like in network/physical service, in that case the starter is defined to be "transient"[00:18:26] <sergefonville1> but how does SMF know whether it is running then?[00:18:59] <tsoome> in case of script? from the fact the start method did end successful[00:20:45] <sergefonville1> but what if the process dies at some point?[00:21:03] <sergefonville1> smf has no idea?[00:22:26] <tsoome> with scripts there are no processes to die[00:22:32] <tsoome> thats the point;)[00:22:38] <tsoome> nothing to watch for[00:26:25] *** synegy34 has joined #openindiana[00:27:57] <sergefonville1> so there is no way SMF can monitor daemonized processes?[00:29:04] <tsoome> there are plenty of daemons in your system:P[00:29:16] <sergefonville1> hahaha[00:29:28] <sergefonville1> I mean the ones that do this[00:29:32] <Triskelios> sergefonville1: if the process dies, SMF will notice[00:29:42] <sergefonville1> yes, that I understand[00:29:52] <sergefonville1> but how do I 'solve' this[00:30:06] <tsoome> for example - autofs service is running automountd daemon[00:30:21] <tsoome> so you can check its manifest and compare[00:30:47] <Triskelios> it should inherit the contract even if it changes sids[00:31:35] <sergefonville1> it gets weirder, if I tell it to daemonize when it isn't started with spawn-fcgi, it works as well[00:33:48] <DanaG> Under SMF, is there any specific benefit to NOT daemonizing?[00:34:57] <DanaG> Like, does it make the logs go somewhere special?[00:36:29] <tsoome> the logs smf is capturing are from start/stop method standard output (and stderr i guess)[00:36:51] <tsoome> any application specific log will go where application wants to put it[00:37:54] <Triskelios> right, the output can be part of the normal service log, and makes it possible to check return values[00:39:39] *** Transformer has joined #openindiana[00:39:55] *** Transformer has left #openindiana[00:41:17] <sergefonville1> when I start a process that daemonizes with spawn-fcgi. is how can that be different from daemonizing without it[00:41:39] <sergefonville1> is the contract lost somewhere?[00:41:56] <sergefonville1> for example when daemonizing a daemonized process?[00:42:07] <sergefonville1> if that even exists....[00:42:30] *** TomJ has joined #openindiana[00:43:45] *** nonot1 has quit IRC[00:44:07] <tsoome> hm, well, its possible its creating new contract for some reason, and i have no idea what will happen then... i think the smf does monitor only initial contract ....[00:46:26] <sergefonville1> I looked at the code for spawn-fcgi and it does do a fork and a setsid ont the process it created[00:48:32] *** gea__ has quit IRC[00:49:57] <tsoome> when it does not daemonize, what will happen then? spawn-fcgi is run with & ?[00:50:19] <sergefonville1> spawn-fcgi always daemonizes[00:50:31] <sergefonville1> but if uwsgi also daemonizes[00:50:37] <sergefonville1> it basically happes twice[00:52:55] <sergefonville1> spawn-fcgi always exits after it created the process[00:54:58] <tsoome> well twice shouldnt be the issue, the whole idea of contracts is that you get subtree or processes, and can monitor its state - catch the faults etc (man ctrun, ctstat, ctwatch). and you can manage how smf can react - ignore some signals, or core or not...[00:55:14] *** SH0x has quit IRC[00:55:53] *** SH0x has joined #openindiana[00:56:39] <sergefonville1> when I set uwsgi to daemonizes and run it with spawn-fcgi smf says it exited[00:56:45] <sergefonville1> if I empty the stop command[00:56:53] <sergefonville1> the processes are still running[01:07:03] <sergefonville1> it creates a copy of the filedescriptot and then closes the old oen[01:07:05] <sergefonville1> one*[01:07:49] *** dijenerate has joined #openindiana[01:10:51] <tsoome> well, what will happen if you dont demonize it?:P[01:11:00] *** gervystar has quit IRC[01:11:38] <tsoome> as its doint there something which disturbs smf, just let it not to demonize and thats it?[01:11:39] <DanaG> argh, why is it so dang hard to find a complete Mono?[01:12:55] <DanaG> One build is missing VB, the other is missing UIAutomationWinforms.exe![01:18:08] <sergefonville1> if I tell spawn-fcgi to not daemonize uwsgi tells me that stdin is not a socket and exits[01:18:51] <sergefonville1> I need spawn-fcgi so I can change the permissions on the socket[01:19:08] <sergefonville1> and run it as a non-privileged user[01:21:40] <sergefonville1> uwsgi creates the socket after dropping priviliges, and since it cannot write in that directory....[01:22:23] <tsoome> which one?[01:22:32] <sergefonville1> /var/run[01:23:38] *** xmikus01 has quit IRC[01:23:53] <tsoome> uhm, well, /var/un belongs ro root and is 755; but create subdir; chown it and create socket there?[01:28:36] <sergefonville1> and I need uwsgi to daemonize, since otherwise it doesn't write to a log file[01:28:50] <sergefonville1> instead it writes to stderr, which doesn't go anywhere[01:29:49] <sergefonville1> I still don't quite understand why daemonizing twice would make it loose the contract[01:29:58] *** InTheWings has quit IRC[01:32:03] <tsoome> no idea what it doest there exactly, cant comment. seems like an case of poorly ported software, seems it needs a bit of fixing - the socket in /var/run, etc...[01:33:14] <sergefonville1> the socket in /var/run was my 'choice'[01:33:31] <sergefonville1> the website says its compatible with oi[01:33:37] <sergefonville1> but there is no mention of smf[01:36:20] *** kovert has quit IRC[01:37:13] <sergefonville1> and it does 'run' on oi[01:37:23] <sergefonville1> just not the way one would expect it to[01:37:47] <sergefonville1> I have a lot of new information to work with[01:37:53] <sergefonville1> time for me to go to sleep[01:38:20] <sergefonville1> thank you very, very much for your guidance and help tsoome[01:38:40] <sergefonville1> and the others that helped as well ofcourse ;-)[01:38:48] <tsoome> well, there are some examples about socket dir etc - they assume the directory is writeable for the socket owner[01:39:59] <sergefonville1> I'll look into that[01:40:10] <tsoome> the /var/run in OI is ok, because its there, but you need to specify subdir there (i dont know if that spawn-fcgi will create it or not)[01:40:43] <sergefonville1> spawn-fcgi creates the file whereever I tell it to, same for uwsgi[01:41:16] <sergefonville1> the reason for using /var/run is that is what I used for php and it made 'sense' to place socket files there[01:41:32] <tsoome> ye, thats ok[01:41:51] <sergefonville1> I also read something about an abstract socket, but could not find if oi supports that[01:42:25] <tsoome> http://redmine.lighttpd.net/wiki/lighttpd/HowToSetupFastCgiIndividualPermissions step 4 - they create startup script to prepare system for that spawn-fcgi[01:42:26] <sergefonville1> nor do I know if nginx does[01:44:00] *** Naresh` has joined #openindiana[01:44:25] <tsoome> http://redmine.lighttpd.net/wiki/spawn-fcgi/Basic_Ideas there they just give socket location n /var/run/subdir,[01:44:31] <tsoome> s/n/in/[01:44:43] <sergefonville1> that looks a lot like what I did for php[01:44:52] <sergefonville1> but php-cgi does not daemonize[01:45:17] <tsoome> but again its good question if the spawn will create that subdir if its not there and ser owner, or if it does expect it there already[01:45:41] *** Naresh has quit IRC[01:45:55] *** melliott has quit IRC[01:46:02] <tsoome> if its expecting it to exist, your service start script needs to create it and set owner[01:46:03] <sergefonville1> the problem is not if spawn-fcgi creates the socket properly, because it does[01:46:23] <sergefonville1> the problem is the double daemonzing on one end[01:46:45] <sergefonville1> or uwsgi not being ablo to create the socket with the right permisions on the other end[01:46:55] <sergefonville1> if I run uwsgi as root[01:47:01] <sergefonville1> no problem[01:47:04] <tsoome> yea, well, that was one issue you had. checking out whats wrong with demonize is a bit more complicated and its way too late for that:)[01:47:19] <sergefonville1> it is very late :P[01:47:26] <sergefonville1> here its 2 AM :P[01:48:23] <sergefonville1> I should;ve gone to sleep three hours ago...[01:48:34] <sergefonville1> so now is as good as ever :P[01:48:48] <DanaG> say, how do I force permissions of all files in some dir, to be inherited from the dir above?[01:49:05] <sergefonville1> use the chmod that came with oi[01:49:16] <sergefonville1> there are extended acls that do that[01:49:30] <sergefonville1> but, I'm off, more tomorrow ;-)[01:49:37] <sergefonville1> thanks a lot[01:49:38] <sergefonville1> bye[01:49:40] *** sergefonville1 has left #openindiana[01:50:00] <DanaG> man chmod on;ly shows gnu chmod.[01:50:57] <tsoome> man /bin/chmod :P[01:51:37] <tsoome> http://mattwilson.org/blog/solaris/solaris-cifs-server-and-zfs-acls-the-problem/[01:51:48] <tsoome> read that blog entry and second comment[01:54:29] <DanaG> How do I change nothing but inheritance flag?[01:54:53] <DanaG> I want to take each of the ACL objects, and make it inheritable.[01:55:54] *** ivo_ has quit IRC[01:55:57] *** DrLou has quit IRC[01:59:25] <DanaG> ah, /bin/ls -V, and copy and paste.[01:59:30] <DanaG> That's fine for me.[02:00:33] *** darkscrypt has joined #openindiana[02:00:45] <darkscrypt> so this might be an easy question, but[02:01:02] <dkeav> the answer is yes[02:01:04] <darkscrypt> is there any way to make a burnable CD with a few packages for solaris, and install them with PKG install?[02:01:21] <darkscrypt> instead of having to download them off the internet?[02:01:29] <dkeav> hah, i'm psychic[02:01:40] <darkscrypt> thats a cool trait to have.[02:02:08] <dkeav> man mkisofs[02:02:41] <darkscrypt> no no, i know how to make a CD. i typically use genisofs, but basically the same thing[02:02:41] <tsoome> well, afaik the ipkg on disk format is not yet there, or im wrong?:P[02:03:09] <dkeav> oh you want ips packages locally[02:03:23] <darkscrypt> like i only want 3 packages, but yeah, I want to burn them to a disk[02:03:36] <darkscrypt> and point to the disk for installs[02:03:50] <tsoome> however, its possible to build local package cache, so you dont have to get every one from net...[02:05:05] <darkscrypt> could that cache be on a CD Rom, that could be used by multiple vanilla (fresh install) OI boxes?[02:06:16] <alanc> tsoome: it's there in pkg sources, but recently enough that I don't think OI builds all have it yet[02:06:49] <dkeav> darkscrypt: so yes it can be done, just not quite yet, unfortunatly for you[02:06:49] <alanc> could still pkgrepo create a new repository and pkgrecv those packages there[02:07:14] <alanc> all the on disk format gives you is a single archive file instead of a big directory tree of files[02:07:26] <dkeav> yea if you have a bunch of OI boxes on a network, just setup a repo of your own on a laptop or something and take to that network[02:07:29] <dkeav> easy peasy[02:07:43] <darkscrypt> hmm.[02:07:51] <alanc> or on a cd/disk - don't need to run a repo server for it, just have filesystem access[02:08:08] <darkscrypt> alanc, that looks like the best[02:08:13] <dkeav> aye for one-at-atiming it[02:08:20] <dkeav> i like to multitask when imaging :D[02:08:49] <darkscrypt> dkeav, its not necessairly an imaging process. many of these installs might not have network access right up front[02:08:59] <darkscrypt> plan on using it at a few locations in hati right now[02:09:36] <darkscrypt> having everything on disk without network access is a must[02:10:08] <darkscrypt> I know how to slipstream files onto the install disk and still have the installation process go unaffected[02:10:26] <darkscrypt> later I just want the installer to call one perl script that is on that install cd after the install[02:10:34] <darkscrypt> and have it install the packages from the disk itself[02:14:56] *** dijenerate has quit IRC[02:19:01] <DanaG> weird, I tried to set acls to give transmission user read-write, and now it gets permission denied instead.[02:19:04] <DanaG> anyway, gotta' go...[02:23:29] *** DanaG has quit IRC[02:36:55] *** redgone has joined #openindiana[02:43:18] <blues> anyone around wanna walk me through creating my first iSCSI target with napp-it?[02:44:45] *** dijenerate has joined #openindiana[02:48:50] *** GateKeeper has joined #openindiana[02:49:03] *** GateKeeper is now known as Guest15116[02:51:40] *** dijenerate has quit IRC[02:51:50] *** synegy34 has quit IRC[03:12:02] *** master_of_master has quit IRC[03:13:50] *** master_of_master has joined #openindiana[03:48:27] *** dijenerate has joined #openindiana[03:52:31] *** McBofh has quit IRC[04:01:23] *** McBofh has joined #openindiana[04:08:08] *** jkimball4 has quit IRC[04:42:32] *** kart_ has joined #openindiana[04:55:15] <gnomad> installing openindiana b151... chug chug chug...[05:03:28] *** skeeziks has joined #openindiana[05:03:50] *** sivanov_ has quit IRC[05:04:36] *** sivanov_ has joined #openindiana[05:06:47] *** sivanov_ has quit IRC[05:07:47] *** sivanov has joined #openindiana[05:13:52] *** POloser has joined #openindiana[05:48:40] *** jkimball4 has joined #openindiana[05:49:23] *** Reepicheep has quit IRC[05:49:24] *** oninoshiko has quit IRC[05:54:59] *** redgone has quit IRC[06:05:29] <gnomad> is authentication (from the GUI) broken in OI b151?[06:06:09] *** Naresh has joined #openindiana[06:06:14] *** Naresh` has quit IRC[06:06:21] <alanc> you mean the bit about trying to run commands with extra priviledges? isn't that in the release notes?[06:06:45] <alanc> (assuming they copied both the bug and the release note for it from the Oracle b151 release)[06:07:00] <blues> anyone use gmail's smtp server to send notices from OI/ napp-it ?[06:07:13] <gnomad> checking for release notes now...[06:08:29] <richlowe> tsoome, alanc: pkg on-disk format, caching, better zones support have been there for a while now (but not on OI)[06:08:37] <richlowe> unless you run aszeszo's pkg bits[06:09:20] *** samc has quit IRC[06:10:19] <alanc> if I recall correctly, the short answer is root's password is initialized as expired and the gui didn't know how to deal with that until after b151, so you have to open a terminal and su root first to initialize a non-expired password for it[06:11:02] <gnomad> So does OI use a root password or not?[06:11:15] <alanc> it does[06:11:36] <gnomad> and it wants the root password from the GUI, not an admin user's password?[06:13:16] <gnomad> ok. that works.[06:13:50] <gnomad> is that the way things are supposed to work in the long-term forward?[06:14:37] *** tsukasa has quit IRC[06:14:41] <alanc> assuming they keep copying those bits from Oracle, the GUI prompts become clearer about which password is requested, and sudo is set up as an option as well[06:14:47] *** Naresh has quit IRC[06:15:14] <alanc> I don't know if OI has a more specific plan or not[06:15:41] *** synegy34 has joined #openindiana[06:16:18] <richlowe> hahahaha[06:16:21] <richlowe> "specific plan"[06:18:37] <richlowe> they have subject tags to argue about, apparently.[06:18:41] *** Guest15116 has quit IRC[06:18:41] <alanc> "The OpenSolaris distros were created by Sun. They survived. They evolved. There are many copies. And they are Creating Plan..../"[06:18:43] <gnomad> I must say, I have gotten horribly spoiled by OS X.[06:19:02] * alanc recycles a punchline from another channel[06:19:07] <joffe> alanc[06:19:09] <joffe> i love that.[06:19:17] <joffe> now i want a Sun box with a cylon visor[06:19:22] <joffe> and that sound[06:19:29] <gnomad> but alas, there is no ZFS for OS x.[06:19:30] *** kovert has joined #openindiana[06:19:41] <alanc> Sun 3's used to have cylon status lights - back and forth[06:19:45] <skeeziks> gnomad: Have you checked out MacZFS?[06:19:48] <skeeziks> I use it every day.[06:20:14] <gnomad> I was under the impression that was even more ancient than the ZFS in FreeBSD.[06:20:26] <skeeziks> Ancient, but functional.[06:20:35] <richlowe> alanc: and a real diag switch![06:20:43] <alanc> also http://tenscomplement.com/z-410-storage-main-features[06:21:20] <skeeziks> alanc: It's released?!?[06:21:24] <gnomad> The hardware in my ZFS box is hackintosh ready. When the ten-complement ZFS comes out, I'll probably go that route.[06:21:42] <alanc> skeeziks: no, website still says coming soon and "in final stages of development"[06:21:57] <gnomad> but in the meantime, I'll stick with OpenSolaris.[06:21:58] <skeeziks> Ah, okay then. It's still more than the last time I checked :)[06:22:36] <gnomad> I've been through countless versions at this point.[06:23:05] <gnomad> Indiana, OSOL, NCP3, and now OI.[06:23:22] <skeeziks> Four! I count four.[06:23:22] <skeeziks> ;)[06:23:24] <gnomad> and SXCE before that[06:23:41] <gnomad> and there were a bunch I played with and never got stable[06:23:46] <gnomad> or working with my hardware.[06:23:58] <richlowe> nexenta need to come up with a way to name a successor product to NCP in a way that acronymizes to IP[06:24:14] <skeeziks> NIP[06:25:32] <gnomad> anyways, I really want to like an Open Solaris, but it's just such a FPITA.[06:26:13] <richlowe> not much more so than the closed one[06:26:22] <gnomad> heh[06:26:36] <richlowe> at the very least, it doesn't cost thousands of dollars to yell at us[06:26:51] <gnomad> I liked Solaris much better when it was BSD-based.[06:27:43] <alanc> Yelling at Oracle is free. Having them listen and respond costs money.[06:28:08] <dkeav> +1 internets for you sir for technical correctness[06:29:14] *** tsukasa has joined #openindiana[06:33:13] <gnomad> but anyways, despite my grumblings, the latest beta of OI gives the best initial impression of anything open-solaris I've played with.[06:36:07] <gnomad> although transmission 0.5.11? srsly?[06:36:43] <richlowe> that's a sad artifact of packaging[06:37:20] <alanc> read that as 0.$(uname -r), which was the default version for packages which didn't set one[06:38:12] <gnomad> huh[06:38:42] <gnomad> ok, so it's actually 1.93.... which is still ancient[06:38:53] <dkeav> pkgsrc has 2.22[06:39:01] <dkeav> just sayin[06:39:09] <gnomad> which is still pretty old.[06:40:01] <dkeav> 2.32 is in wip[06:40:32] <gnomad> ok, is there a page that shown how to set-up/use pkgsrc in OI?[06:41:01] <dkeav> https://www.illumos.org/projects/worsoe ??[06:43:05] <gnomad> ok... I'll play with that.[06:43:16] <dkeav> http://pkgsrc-wip.sourceforge.net/[06:43:25] <dkeav> ^^ bleeding edge packages[06:43:30] <dkeav> good luck[06:43:39] <gnomad> getting transmission to compile under NCP was unbelievably painful.[06:44:00] <dkeav> would compile under gcc44 in pkgsrc[06:44:07] <dkeav> after you bootstrap pkgsrc, that is[06:56:18] *** oninoshiko has joined #openindiana[07:12:31] <sivanov> i wonder why my each next iozone test on same FC luns gives me less and less read performance???[07:13:51] *** skeeziks has quit IRC[07:15:27] <sivanov> just bought this motherboard: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=110710885869&ssPageName=ADME:L:OC:US:1123[07:15:36] <sivanov> hope it'll work hehe[07:17:18] <sivanov> does anybody know if i can flash the onboard LSI 1068 into IT mode as people do with discrete cards?[07:17:42] *** syoyo has joined #openindiana[07:17:43] *** hajma has quit IRC[07:17:48] *** miine has quit IRC[07:20:34] *** syoyo has quit IRC[07:21:01] *** syoyo_ has joined #openindiana[07:22:33] *** miine_ has joined #openindiana[07:26:53] *** Naresh has joined #openindiana[07:29:40] *** forquare has joined #openindiana[07:33:30] *** forquare has quit IRC[07:37:39] *** yalu_ has quit IRC[07:39:07] *** yalu has joined #openindiana[07:57:16] *** DanaG1 has joined #openindiana[07:57:40] <DanaG1> Hmm, I tried to give the "transmission" user permissions to certain folders, but that user is getting "access denied".[07:57:46] *** Botanic has quit IRC[07:57:54] <DanaG1> here's my chmod: sudo /bin/chmod -R A=owner@:rwxpdDaARWcCos:df:allow,user:transmission:rwxp--aARWc--s:df:allow /home/dana/Torrents /home/dana/Music /home/dana/Videos /home/dana/Pictures[08:03:30] *** syoyo_ has quit IRC[08:03:56] *** syoyo has joined #openindiana[08:08:51] *** SH0x has quit IRC[08:26:56] *** dekar has quit IRC[08:28:40] *** flyz has quit IRC[08:28:53] *** flyz has joined #openindiana[08:39:01] <DanaG1> weird, I moved my nic to the other slot, and it's not showing up.[08:42:01] *** Botanic has joined #openindiana[08:42:01] *** Botanic has joined #openindiana[08:42:37] <DanaG1> quick: how do I make it use that device? I need to make sure the slot is good.[08:43:10] <miine_> DanaG1: boot from a linux cd and see if its there...[08:43:23] <DanaG1> Well, I can boot OpenSolaris, and I know it works in the other slot...[08:43:33] <DanaG1> er, OpenIndiana.[08:44:28] <miine_> DanaG1: may be the wrong answer here, but at at least linux hw support is much broader... . so if it is not there you can assume it's an hw issue...[08:44:35] <DanaG1> Okay, it worked on second boot.[08:44:47] <DanaG1> It's odd... when I re-enabled bge0, e1000e0 went away.[08:45:04] <nettezzaumana> moin[08:45:04] <miine_> DanaG1: DHCP?[08:45:36] <nettezzaumana> DanaG1: it's done by ignorant nwam ... it tries to be smarter then you[08:46:05] <DanaG1> But e1000e0 didn't even appear in ifconfig -a[08:46:33] <nettezzaumana> DanaG1: ?unplumbed[08:46:38] <miine_> DanaG1: nwam disabled it. I don't think that two dhcp addresses are what you really want...[08:47:22] <miine_> DanaG1: one the same subnet of course...[08:47:27] <nettezzaumana> btw, waz e1000*e*X ?? typo or does this really exist ?[08:48:02] * nettezzaumana knows only e1000*g*X[08:48:12] <DanaG1> That's thinko... going off of memory.[08:48:20] <DanaG1> And Linux driver is e1000e.[08:48:39] <nettezzaumana> what an awkward statement !![08:49:25] <nettezzaumana> DanaG1: oh really, my apologies .. it exists[08:50:07] * DanaG1 wonders how useful it would be to have dd-wrt link-aggregate ports 3 and 4, and have OI aggregate Intel and Broadcom.[08:50:13] <DanaG1> It'd probably be counterproductive.[08:50:26] <DanaG1> Okay, so it is e1000g1[08:50:37] <miine_> DanaG1: dd-wrt on which hw?[08:50:49] <DanaG1> wrt350n.[08:51:01] <DanaG1> It's just acting as an AP and a switch right now.[08:51:02] <nettezzaumana> e1000g & bge[08:51:08] <miine_> DanaG1: and your downlink/uplink is?[08:51:13] <DanaG1> I have tomatousb nousb-vpn on a WHR-g125 as primary router.[08:51:22] <DanaG1> I'm thinking LAN, not WAN.[08:51:33] <DanaG1> All other links would be single GbE.[08:51:43] <DanaG1> Well, except the link to the router, which is 100 megabit.[08:52:18] <DanaG1> But all devices are hanging off a gigabit switch and/or that dd-wrt. Only one link from router to one of the switches.[08:52:25] <miine_> DanaG1: buy a real switch. I wouldn't trust neither dd-wrt on aggregation nor those cheap hw...[08:53:12] <DanaG1> The switch in the other room is one of the Level1 ones from newegg.[08:53:14] <miine_> DanaG1: you would be better of by making two subnets (and routing between) or three subnets...[08:54:13] <DanaG1> Right now I have a sort of fuzzy stick-figure. Router is the head. Big switch is the other end of that link to the router. There are 2 computers and 1 network printer hanging off that switch. There's another link down to my room, the dd-wrt.[08:54:30] <miine_> DanaG1: but anyway - I bet you won't saturate even 1Gb in normal everyday use...[08:54:34] <DanaG1> Microserver is in my room, as well as a printer and 1-2 laptops.[08:54:44] <DanaG1> And bge CPU usage will be evil. =P[08:57:12] <miine_> DanaG1: I've an atom 330 with an realtek 1 Gb nic here. having 3-4 computers accessing it (filesharing, backups, mail, voip, video-streaming (sd) ). No speed problems... . I would stick with the single link OR you have to spend real money on the solution (new switch etc.)[08:57:58] *** bens1 has joined #openindiana[08:58:37] <DanaG1> hmm, what's a good way to test the HDD+network speed?[08:59:29] <DanaG1> I was using "rich copy", and that app was taking 100% cpu on my laptop, so that's not a good test.[08:59:32] <miine_> DanaG1: just test what how you want to use it in reality. otherwise your results are synthetic...[08:59:33] <DanaG1> it's a GUI robocopy.[09:00:25] <DanaG1> oh, and all my permissions are funky, since I tried to use ACLs.[09:00:40] <DanaG1> What I want: only myself and the "transmission" user can read to and write from my user dirs.[09:01:24] <miine_> DanaG1: if you can model access rights with users & groups, just use those...[09:01:55] <miine_> DanaG1: there is no law saying that you have to use every shiny feature available ;-)[09:01:57] *** baitisj has joined #openindiana[09:02:57] *** xmikus01 has joined #openindiana[09:03:15] *** SH0x has joined #openindiana[09:03:16] <DanaG1> okay, so I'd just make a group for me and Transmission?[09:03:28] <DanaG1> Or even easier: a transmission group, and add myself.[09:03:39] <DanaG1> Too bad usermod -G can't append.[09:06:35] <nettezzaumana> DanaG1: usermod -G $foo,`${some brainfuck with groups or id}` $user[09:06:45] *** anikin has joined #openindiana[09:07:06] *** |AbsyntH| has joined #openindiana[09:07:43] *** sivanov has quit IRC[09:07:54] *** Micr0mega has joined #openindiana[09:13:53] <DanaG1> hmm, still permission denied.[09:14:14] <DanaG1> oh, my /home itself.[09:14:50] <DanaG1> Cool. go+x on my home dir.[09:17:54] <DanaG1> argh, it's creating its config file as rw transmission user, only![09:19:03] *** Whoopsie has joined #openindiana[09:19:03] *** ChanServ sets mode: +v Whoopsie[09:28:11] *** dekar has joined #openindiana[09:31:58] *** lkocman has joined #openindiana[09:31:58] *** lkocman_ has joined #openindiana[09:32:45] *** syoyo has quit IRC[09:37:14] <DanaG1> okay, fixed my read permissions, but it's still saying I can't write to some of my files![09:39:42] <DanaG1> argh, some of my filenames have turned into this:[09:39:42] <DanaG1> 謫社ＣＤã³ããã¯ãã«ã¼ããã£ãã¿ã¼ããã[09:39:54] <DanaG1> That filename was originally kanji, I believe.[09:40:12] *** tolkor has quit IRC[09:40:25] <tsoome> heh...[09:41:08] <lblume> Locale badly set?[09:42:23] <DanaG1> Not sure.[09:42:45] <DanaG1> It was (and still is) fine on the source, in utf-8.[09:42:57] <DanaG1> I'd assume it's not using utf-8. Should I just make it utf8only?[09:43:44] *** tolkor has joined #openindiana[09:55:22] *** miine_ has quit IRC[09:56:57] *** syoyo has joined #openindiana[09:56:59] *** miine has joined #openindiana[10:02:24] *** miine has quit IRC[10:11:41] *** sivanov has joined #openindiana[10:11:49] *** sivanov is now known as sivanov_[10:13:31] *** pjama has quit IRC[10:13:47] <RoyK> hm... how can I find a drive's serial number from OI?[10:15:17] <tsoome> iostat -E does not list it?[10:15:34] <tsoome> like Vendor: SEAGATE Product: ST314670FSUN146G Revision: 055A Serial No: 061944DFY7[10:16:05] <DanaG1> my microserver came with a drive that calls itself a (V)B(250)EAVER[10:16:09] <DanaG1> Parentheses are added by me.[10:16:44] <lblume> If iostat does not show it, try smartctl[10:20:31] *** Lumb_ has joined #openindiana[10:20:50] *** pettson_ has joined #openindiana[10:21:25] *** robinbowes has quit IRC[10:22:01] <DanaG1> say, is there an equivalent of lm-sensors?[10:22:17] * DanaG1 tries pkg search sensors[10:22:17] <RoyK> tsoome: http://pastie.org/2200869[10:23:04] <tsoome> ye got similar result on another host[10:23:36] *** heldchen has joined #openindiana[10:23:49] *** Gugge_ has joined #openindiana[10:25:16] *** nikolam has joined #openindiana[10:25:33] *** Whoopsie has quit IRC[10:25:33] *** flyz has quit IRC[10:25:33] *** POloser has quit IRC[10:25:33] *** held has quit IRC[10:25:35] *** ianj has quit IRC[10:25:35] *** Hedonista has quit IRC[10:25:35] *** LizardKing has quit IRC[10:25:35] *** Gugge has quit IRC[10:25:35] *** chototsu has quit IRC[10:25:36] *** Lumb has quit IRC[10:25:37] *** pettson has quit IRC[10:25:37] *** pettson_ is now known as pettson[10:26:08] *** flyz has joined #openindiana[10:26:36] *** LizardKing has joined #openindiana[10:27:17] <tsoome> T5220 with sas does report serials.... so its depending on disk or hba...[10:30:08] <lblume> Or combination of both. iostat does not try very hard.[10:31:03] <DanaG1> what's the equivalent of lspci?[10:33:45] *** TibshoOT has joined #openindiana[10:36:51] <lblume> scanpci/prtconf[10:37:08] *** Norrin has quit IRC[10:40:17] *** robinbowes has joined #openindiana[10:40:59] <tsoome> iostat doesnt have to - it only does display what is recorded in kstat[10:41:03] *** dekar has quit IRC[10:42:34] <lblume> Or combination of both. kstat does not try very hard.[10:42:48] *** DanaG1 has quit IRC[10:43:28] <tsoome> kstat does not digg anything either;) all data is fed into it:P[10:44:19] *** oninoshiko has quit IRC[10:44:28] <tsoome> sd/sata does not get it from its looking. (ye, probably looking from not too many places:)[10:45:22] <lblume> Lazy thing :-)[10:46:44] <nikolam> What are other IPS package repositories that you know ad that are working?[10:47:38] <nikolam> Btw, installing Virtualbox on /dev-il 151 needs reboot after uninstalling previous Vbox.. Unlike before.[10:54:33] *** InTheWings has joined #openindiana[10:57:46] <lblume> the modules did not unload?[11:14:23] *** freedomrun has joined #openindiana[11:17:43] *** man_u has quit IRC[11:27:39] *** McBofh has quit IRC[11:35:55] *** heldchen has quit IRC[11:35:57] <nikolam> lblume, uh, I restarted it and reinstalled it. I would need to go back to previous release install to see actual message.[11:36:18] <nikolam> Should such things be reported as bugs or reported to upstream/Vbox[11:38:36] <lblume> *shrug* don't bother[11:38:44] *** gervystar has joined #openindiana[11:39:05] <lblume> I have not been able to update VBox without rebooting since v 3.0 or something.[11:39:47] <lblume> I opened a bug, and asked on #vbox, and met very little interest. It boils down to «It works for me»[11:39:52] <nikolam> you repported it, and talked about it on mailing list? I think I did update vbox without reboot, untill I got in /dev-il[11:40:47] <nikolam> I am afraid of they continuing with Solaris closed and not bothering with Illumos/OI compatibility of there is no activity on this side[11:41:13] <nikolam> Like compiling own Vbox free and putting it in IPS etc.[11:42:41] <lblume> Well, worth asking, at least.[11:43:56] <nikolam> lblume, I just wonder you say you could not uninstall/reinstall without rebooting, But I think I could at least for some previous releases before this[11:46:32] <lblume> Yes, but unfortunately, at the time it happened, I was not physically close to the box and was not able to diagnose, and then I just needed it to work.[11:47:09] <lblume> And at first, I believed I was doing something wrong (I had some fun experience in the past with their scripts)[11:47:39] <lblume> But now, there is no doubt, none of their module will unload, under any condition.[11:50:08] <nikolam> aha I see. And they probably DO unload on Solaris11 express or something.. hmm[11:50:23] <nikolam> I could check that out, since I have S11Ex here in BE.[11:51:15] *** POloser has joined #openindiana[11:57:28] *** held has joined #openindiana[11:58:08] *** lkocman has quit IRC[11:58:28] *** lkocman has joined #openindiana[12:02:28] <lblume> nikolam: It should work. There is definitely something fishy with my system, just I have not been able to find out what, and did not get much help.[12:02:51] *** McBofh has joined #openindiana[12:03:47] <lblume> Oh, I thought I had said, that's not even on OI, it's on an S10 box, so it should be really 100% supported.[12:06:41] <nikolam> ah, that is another issue then.[12:07:23] *** hajma has joined #openindiana[12:08:25] *** bens1 has quit IRC[12:10:12] <DontKnwMuch> royk: smartctl -a -d sat,0 /dev/rdsk/c3t5000CCA225C1EA7Cd0 shows me serials, even when nothing else does[12:10:31] <lblume> nikolam: Probably, but if you do find a solution to yours, I'd like to know how you did[12:12:22] <nikolam> lblume, is there a bug yet, you reported for unloading modules after uninstall of Vbox on S10? What Update was that, 8?[12:15:22] *** dekar has joined #openindiana[12:16:24] *** |AbsyntH| has quit IRC[12:18:05] *** LeftWing has quit IRC[12:18:11] *** LeftWing has joined #openindiana[12:18:57] <lblume> the bug is there: http://www.virtualbox.org/ticket/7799[12:19:27] *** dekar has quit IRC[12:25:00] *** Hedonista has joined #openindiana[12:25:01] *** ianj has joined #openindiana[12:25:05] *** chototsu has joined #openindiana[12:25:05] *** ChanServ sets mode: +v chototsu[12:31:14] *** andygraybeal has joined #openindiana[12:42:03] *** jedi4ever has joined #openindiana[12:49:09] <sivanov_> Children see throughput for 2 random readers = 7544.82 KB/sec[12:49:26] <sivanov_> Children see throughput for 2 random writers = 202770.48 KB/sec[12:49:49] <sivanov_> just dont understand whats up with my LUNs[12:55:21] *** POloser has left #openindiana[13:01:55] *** DrLou has joined #openindiana[13:01:55] *** ChanServ sets mode: +o DrLou[13:02:31] <sivanov_> each time i run iozone, the read speed from LUN declines[13:08:14] *** bens1 has joined #openindiana[13:12:21] <lennard> hmm, I don't think my oi-dev-148-x86.usb wants to boot[13:12:35] <lennard> OR i just download the -for-dd file[13:12:44] <lennard> nvm me[13:14:52] *** dekar has joined #openindiana[13:17:02] *** syoyo___ has joined #openindiana[13:20:34] *** syoyo has quit IRC[13:50:43] <RoyK> anyone here that has a 4k sector drive?[13:51:24] <DeanoC> yep[13:51:42] *** laserbled has joined #openindiana[13:51:55] *** sharks has joined #openindiana[13:52:03] *** laserbled has quit IRC[13:52:06] <DeanoC> well in a machine 100+ miles away[13:52:17] *** jedi4ever has quit IRC[13:53:12] *** syoyo___ has quit IRC[13:56:32] *** sharks has quit IRC[13:57:03] *** syoyo_ has joined #openindiana[13:57:15] <RoyK> DeanoC: can you pastebin smartctl -i -d sat,0 /dev/rdsk/devname for such a drive[13:57:19] <RoyK> ?[13:57:28] *** magyar has quit IRC[13:58:48] <nikolam> Is there normal tutorial for installin OI on top of ZFS for such drives/pools ?[13:59:08] <tsoome> such drives?:P[13:59:47] <lennard> 4k sector drives[14:00:27] <tsoome> only real "issue" can be the ones which do lie[14:00:49] <DeanoC> RoyK: sure one sec[14:01:00] <DeanoC> theres is a page on the wiki about them nikolam[14:01:22] <tsoome> rest is just an consideration - do you really wanna have rpool on some 2-3 TB or not.[14:01:35] <RoyK> DeanoC: writing a small parser to associate serialnumber/wwn/devname[14:03:40] <DeanoC> RoyK: don't suppose you know which pkg smartctl in, don't seem to have installed[14:04:29] <tsoome> you can compile it with ./configure;make[14:06:54] <DeanoC> okay i'll build on my build machine and when do get you the results[14:08:41] *** syoyo_ has quit IRC[14:18:02] <DeanoC> RoyK: Its jsut giving me device is busy and I can't really take it off line for a week or two, sorry :([14:25:02] <nikolam> thanks DeanoC[14:30:50] *** kdavy has quit IRC[14:30:59] <RoyK> DeanoC: you need to compile it from source[14:31:14] <RoyK> DeanoC: what sort of controller?[14:34:49] <DeanoC> yeah compiled from source. ATA its reporting as if I do a ./smartctl --scan-open (though its a SATA)[14:34:57] <DeanoC> just can't get any other info[14:35:19] <DeanoC> though one error message is "ATA command routine ata_command_interface() NOT IMPLEMENTED under Solaris. "[14:35:29] *** mikw has joined #openindiana[14:38:34] <RoyK> hm[14:39:49] <RoyK> but what sort of controller?[14:41:59] <lblume> tsoome: You can't have rpool on a > 2TB volume[14:42:17] <lblume> DeanoC: Try to force smartctl to use scsi[14:43:24] <DeanoC> nope scsi no different[14:45:24] <DeanoC> RoyK: SI3114 I beleive[14:45:32] <lblume> aaahhh[14:46:00] <lblume> Running in IDE mode? cXdY devices, not cXtYdZ?[14:46:06] <DeanoC> yep[14:46:19] <lblume> smartctl can't work with that[14:46:39] <lblume> You need scsi/sas/sata[14:47:48] *** DontKnwMuch has quit IRC[14:47:55] *** miine has joined #openindiana[14:49:02] <tsoome> aye, up to 2TB is supported....[14:53:20] *** kimc has quit IRC[14:55:55] *** |AbsyntH| has joined #openindiana[15:07:01] *** nikolam has quit IRC[15:08:32] *** nikolam has joined #openindiana[15:15:55] <RoyK> DeanoC: oh - now that's a PoS controller :P[15:24:37] <DeanoC> RoyK: you leave the old girl alone! ageist! she's going into retirement in the next few weeks, after a hard life :P[15:38:43] *** movement has quit IRC[15:51:05] *** movement has joined #openindiana[15:53:14] *** miine has left #openindiana[15:53:45] *** miine has joined #openindiana[15:53:50] *** Guest4340 has joined #openindiana[15:58:15] *** Guest4340 has left #openindiana[16:02:30] *** dekar has quit IRC[16:14:37] *** nikolam has quit IRC[16:40:39] *** Micr0mega has left #openindiana[16:45:28] *** dekar has joined #openindiana[16:50:54] *** adama has joined #openindiana[16:52:23] *** anikin has quit IRC[16:52:53] *** anikin has joined #openindiana[16:54:35] *** Gugge_ is now known as Gugge[16:56:39] *** Naresh has quit IRC[16:56:41] *** syoyo_ has joined #openindiana[16:57:17] *** anikin has quit IRC[17:15:42] *** hsp has joined #openindiana[17:23:50] *** DanaG has joined #openindiana[17:25:43] *** miine has quit IRC[17:32:20] *** miine has joined #openindiana[17:37:39] *** tsoome has quit IRC[17:38:50] *** scarcry has joined #openindiana[17:39:11] *** scarcry has quit IRC[17:39:42] *** xmikus01 has quit IRC[17:40:02] *** scarcry has joined #openindiana[17:40:07] *** scarcry has quit IRC[17:40:22] *** lkocman_ has quit IRC[17:40:22] *** lkocman has quit IRC[17:40:24] *** freedomrun has quit IRC[17:40:45] *** scarcry has joined #openindiana[17:44:24] *** tsoome has joined #openindiana[17:45:55] *** Orionau has joined #openindiana[17:45:56] *** Oriona has quit IRC[17:51:15] *** kdavy_ has joined #openindiana[17:51:40] *** Orionau is now known as Oriona[17:52:35] *** hajma has quit IRC[17:56:57] *** tsukasa has quit IRC[17:57:49] *** Reepicheep has joined #openindiana[18:02:41] *** Oriona has quit IRC[18:02:52] *** Oriona has joined #openindiana[18:07:59] *** bradend has joined #openindiana[18:22:25] *** skeeziks has joined #openindiana[18:23:32] *** trn has quit IRC[18:24:33] *** flyz has quit IRC[18:26:22] *** flyz has joined #openindiana[18:29:04] *** held has quit IRC[18:29:15] *** tsukasa has joined #openindiana[18:34:54] *** syoyo_ has quit IRC[18:35:48] *** |AbsyntH| has quit IRC[18:44:39] *** Oriona has quit IRC[18:46:11] *** TomJ has quit IRC[18:48:40] *** mikw has quit IRC[18:52:30] *** melliott has joined #openindiana[18:52:43] *** trn has joined #openindiana[19:04:21] *** trn has quit IRC[19:05:20] *** sergefonville has joined #openindiana[19:05:32] *** TomJ has joined #openindiana[19:05:35] *** sergefonville has left #openindiana[19:05:39] *** sergefonville has joined #openindiana[19:05:53] <sergefonville> good evening[19:06:11] <sergefonville> how is it here?[19:06:43] <tomww> hot[19:07:30] *** echobinary has quit IRC[19:07:58] <DanaG> interesting: people in #ubuntu-server are saying that the broadcom 5723 is better hardware than the Intel 82574.[19:10:05] <sergefonville> its ububt...[19:10:15] <sergefonville> ubuntu*[19:10:23] <sergefonville> what more is there to say :p[19:10:29] <tsoome> what makes it better?[19:15:01] <DanaG> Beats me.[19:15:07] <DanaG> I don't know either way.[19:15:29] <tsoome> i see:)[19:15:55] <DanaG> Something about that intel card being a budget card.[19:16:09] <DanaG> I guess the real way to figure out what's "best" is to try both.[19:16:27] <DanaG> Intel is 82574L. Broadcom is 5723.[19:16:47] <DanaG> Oh yeah, now for a different topic: where's something like lm_sensors for opensolaris?[19:16:55] *** Naresh has joined #openindiana[19:20:20] <tomww> normally for supported hardware there is platform support builtin, or here is a servicecontroller able to talk to the system management bus[19:26:16] <sergefonville> in case anyone cares... :P[19:26:24] <sergefonville> I asked on the uwsgi channel[19:26:38] <sergefonville> and the feature I requested is added :D[19:26:58] <DanaG> On my server, there are some smbus temperature probes, as well as a BMC that specifies the wrong base address (and thus the driver won't load).[19:27:14] <DanaG> smbus probes include my RAM. There's a sensor on one stick.[19:31:30] *** gea has joined #openindiana[19:31:36] *** kdavy_ has quit IRC[19:35:05] *** hajma has joined #openindiana[19:44:46] *** hajma has quit IRC[19:46:34] *** Naresh` has joined #openindiana[19:47:33] *** forquare has joined #openindiana[19:48:13] *** Naresh has quit IRC[19:48:35] *** piwi has joined #openindiana[19:55:31] *** Botanic has quit IRC[19:58:56] *** echobinary has joined #openindiana[20:02:36] *** ivo_ has joined #openindiana[20:08:14] *** kart_ has quit IRC[20:09:55] *** gervystar has quit IRC[20:10:22] *** kytibe has quit IRC[20:12:24] *** gber has joined #openindiana[20:15:46] *** kytibe has joined #openindiana[20:16:33] *** Botanic has joined #openindiana[20:16:33] *** Botanic has quit IRC[20:16:33] *** Botanic has joined #openindiana[20:25:32] *** bens1 has quit IRC[20:41:40] *** russiane39 has joined #openindiana[20:54:09] *** Naresh`` has joined #openindiana[20:54:52] *** AlasAway is now known as Alasdairrr[20:55:52] *** Naresh` has quit IRC[21:16:16] *** gber_ has joined #openindiana[21:16:31] *** wonslung has quit IRC[21:19:33] *** gber has quit IRC[21:19:33] *** gber_ is now known as gber[21:29:29] *** smrt has quit IRC[21:29:48] *** smrt has joined #openindiana[21:31:36] *** bens1 has joined #openindiana[21:32:06] *** dijenerate has quit IRC[21:32:16] *** forquare has quit IRC[21:34:56] *** Naresh`` is now known as Naresh[21:34:59] *** Naresh has joined #openindiana[21:43:52] *** wonslung has joined #openindiana[21:46:55] *** piwi has quit IRC[22:10:19] *** held has joined #openindiana[22:21:48] *** SH0x has quit IRC[22:25:43] *** bens1 has quit IRC[22:57:26] <Hedonista> brushing your teeth instead of having a smoke really really helps kill that craving[22:57:43] <Hedonista> ww[22:58:10] *** Lumb_ is now known as Lumb[23:00:06] *** SH0x has joined #openindiana[23:02:01] *** bradend has quit IRC[23:09:23] <richlowe> also, makes your soda taste like ass too[23:09:30] <richlowe> so you still get that "mouth tastes like a portajohn" side-effect.[23:15:19] *** bradend has joined #openindiana[23:22:08] <jkimball4> mmmm[23:22:36] <jkimball4> now if it were hyacinth flavored toothpaste you'd be set[23:23:10] *** raichoo has joined #openindiana[23:26:13] <Lumb> hmm[23:26:41] <Hedonista> honeysuckle would be a good flavor[23:32:23] *** gber has quit IRC[23:40:29] *** raichoo has quit IRC