[00:05:36] <PatrickS> ups, just been away from the keyboard...
[00:07:29] <PatrickS> tank is a raidz2 of 6 x 0.75TB
[00:08:19] <PatrickS> and the I/O spped to my windows box are roughly
[00:08:28] <PatrickS> 26MB/sec write
[00:08:34] <PatrickS> 30MB/sec read
[00:08:52] <PatrickS> viewed from the windows box
[00:11:40] <mib_chrol> I find xfer speeds vary a lot; vista especially - I get between 5 and 80 MB/s (sometimes a gig takes only 13 secs) (raidz) on my ncp1++ box
[00:12:42] <mib_ol92em> ok
[00:13:33] <mib_ol92em> now I'm trying to install a desktop to Nexenta and I'm getting a broken packages error with both gnome and xfce4
[00:23:38] *** randybias has joined ##nexenta
[00:26:55] *** kim0 has quit IRC
[00:37:41] *** olsner has quit IRC
[00:42:21] <mib_ol92em> I'll come back later, now I ahve to go sleep, good night
[00:42:27] *** mib_ol92em has quit IRC
[00:46:19] *** randybias is now known as randybias|away
[01:09:47] *** PatrickS has quit IRC
[01:21:47] *** mlo has quit IRC
[03:17:03] *** perfectgroup has joined ##nexenta
[03:19:57] <perfectgroup> which is the safest PCI-E gigabit desktop ethernet card to use with Nexenta/OpenSolaris?
[03:22:03] <dtbartle> i would think the intel ones are pretty safe
[03:22:54] <perfectgroup> never knew the Realtek have such problems of dropping the connection :(
[03:23:29] <perfectgroup> seems yet to resolve the issue with the rge driver after so long...
[03:27:58] *** anilg has joined ##nexenta
[03:29:13] *** perfectgroup has quit IRC
[03:30:47] *** anilg has quit IRC
[03:38:44] *** anilg has joined ##nexenta
[03:41:18] <viridari> perfectgroup is gone but if the question comes up again, I've been fine with Intel & nvidia GigE
[03:41:33] *** rootard_zzz is now known as rootard
[03:42:33] <anilg> morning all
[03:42:43] <rootard> anilg: morning
[03:43:21] <viridari> hello anilg, rootard
[03:43:28] *** rootard has left ##nexenta
[03:43:33] *** rootard has joined ##nexenta
[03:43:34] *** ChanServ sets mode: +o rootard
[03:44:15] <anilg> heya viridari, how goes the weekend hacking?
[03:48:39] *** anilg2 has quit IRC
[03:48:49] *** rootard changes topic to "Welcome to the Nexenta IRC Channel |  NexentaCP 2 Beta2 released: http://www.nexenta.org/os | Please coordinate bugs against hardy-testing | NCP2 will be released in 2 weeks so file your bugs and voice your opinion now or forever hold your peace! | https://launchpad.net/nexenta/+milestone/hardy-release | Bug reports: https://bugs.launchpad.net/nexenta/+filebug"
[03:49:04] *** rootard is now known as rootard_zzz
[04:08:43] *** randybias|away is now known as randybias
[04:10:39] *** tsukasa` has joined ##nexenta
[04:25:37] *** tsukasa__ has quit IRC
[04:34:42] *** randybias is now known as randybias|away
[04:54:45] <viridari> howdy anilg ... was doing some much needed decompression via warzone2100 :) still trying to think of how I'd like to go about doing a mini NAS demo, highlighting the cost/performance benefits of HSP
[04:59:39] *** randybias|away is now known as randybias
[06:31:00] *** lesterc has joined ##nexenta
[06:36:15] *** randybias has quit IRC
[06:49:50] *** Corwin has joined ##nexenta
[07:09:10] <Corwin> Is there anything compariable to KVM for Nexenta?
[07:12:05] <dtbartle> virtualbox?
[07:44:33] <Corwin> virtualbox is not in the repositories
[07:44:49] <dtbartle> it's a work-in-progress afaik
[07:45:05] <Corwin> Do you know if it's stable enough for production?
[07:45:11] <dtbartle> you could email the list and ask for details
[07:45:53] <Corwin> Probably be easier to google it... Do you know if you can run virtualbox w/o the GUI crap?
[07:46:28] <dtbartle> i have no idea
[07:46:41] <Corwin> ok, well thanks for the pointer. :)
[08:32:26] *** olsner has joined ##nexenta
[08:58:50] *** kim0 has joined ##nexenta
[09:25:16] *** mib_i3k7ov has joined ##nexenta
[09:42:34] *** mib_i3k7ov has quit IRC
[09:42:34] *** mib_chrol has quit IRC
[09:42:34] *** mib_y2d54y has quit IRC
[09:43:26] *** mib_i3k7ov has joined ##nexenta
[09:44:13] *** mib_y2d54y has joined ##nexenta
[09:45:18] *** mib_chrol has joined ##nexenta
[11:36:20] *** chrol has joined ##nexenta
[11:39:33] *** chrol has quit IRC
[11:41:38] *** chrol has joined ##nexenta
[12:00:36] *** kim0 has quit IRC
[12:00:53] *** kim0 has joined ##nexenta
[12:09:48] *** mib_chrol has quit IRC
[12:22:30] *** mib_i3k7ov has quit IRC
[12:46:07] *** mib_34emay has joined ##nexenta
[13:00:55] *** proteusguy has quit IRC
[13:08:03] *** mib_3j18eo has joined ##nexenta
[13:08:45] *** mib_3j18eo is now known as grntea
[13:14:28] *** grntea has quit IRC
[13:17:44] *** proteusguy has joined ##nexenta
[13:18:44] *** Corwin has quit IRC
[13:32:48] *** mib_34emay has quit IRC
[13:41:09] *** rgl has joined ##nexenta
[13:41:12] <rgl> hi
[13:43:42] <rgl> nexenta 2 will include a desktop?
[13:46:31] <chrol> it will probably be possible to add a desktop but it is not the focus from what i understand
[13:47:01] <chrol> ncp2b2 : I am having issues with passwd; access denied; could someone confirm?
[13:47:56] <rgl> chrol, that was my impressions;  but after reading the beta 2 announcement I've seen it has the qtk and qt libraries too, which left me thinking hehe
[13:48:27] <chrol> Yes - many people have gnome or at least xfce4 running in current beta
[13:50:25] <chrol> false alert: passwd is working fine (i just misspelled pam_smb_passwd.so.1 in pam.conf)
[13:50:46] <chrol> Would it be wrong to have pam_smb_passwd.so.1 installed by default?
[13:57:11] *** uski has joined ##nexenta
[13:57:20] <chrol> In my view ncp2 should be optimized for smb, nfs and iscsi sharing
[13:57:26] <chrol> out of the box
[13:57:32] <uski> hi; will vlc be included in ncp2?
[13:57:41] <uski> i tried to built it a few weeks ago and failed miserably...
[13:58:46] *** mlo has joined ##nexenta
[14:06:02] *** kim0 has quit IRC
[14:29:00] *** lesterc has quit IRC
[14:34:58] *** elbing_ has quit IRC
[14:43:17] *** kim0 has joined ##nexenta
[15:19:58] *** anilg1 has joined ##nexenta
[15:38:01] *** anilg has quit IRC
[15:40:46] *** mib_23salp has joined ##nexenta
[15:43:34] *** mib_23salp has quit IRC
[15:59:29] *** anilg1 has quit IRC
[16:02:05] *** rootard_zzz is now known as rootard
[16:23:14] <mlo> hey rootard, I am making progress on the postfix.  There seems however to be some interaction with nss that causes problems
[16:23:54] <rootard> interesting, what's going on?
[16:24:38] <mlo> Also, whats the general philosophy for solaris specific stuff like /dev/poll ? Should we rather disable it (as is on linux), or provide it (requires /dev/poll in the chroot) ?
[16:25:19] <mlo> Well, it seems that when postfix gets installed, the cache still doesnt know about the groups / userids, so it returns the nobody gid at first.
[16:26:15] <mlo> I've spent the last day trying to track this down by printf debugging in the postfix source without any clear idea why this would happen. Then I finally disabled nscd, and voila, postfix started up.
[16:26:17] <rootard> As long as there is a working mechanism to replace /dev/poll under Solaris then I think it's ok to use that
[16:26:45] <rootard> I ended up bringing /dev/poll into a chroot and called it a day
[16:27:16] <rootard> It sounds like there is a misconfiguration with nss if that is the case
[16:27:29] <rootard> either that or nscd has a new bug
[16:27:45] <mlo> The discussion is a performance discussion. The way to replace dev/poll is using select() / poll(), which is less efficient when having hundreds of connections.
[16:28:57] <rootard> Is there any possible way to select this at runtime?
[16:28:59] *** anilg has joined ##nexenta
[16:29:57] *** anilg has quit IRC
[16:30:11] *** anilg has joined ##nexenta
[16:30:17] <rootard> anilg: just sent you an email
[16:30:29] *** ChanServ sets mode: -o rootard
[16:31:41] <anilg> got it.. never realized I;ve got so much there
[16:31:56] <rootard> yeah, that's why I stopped keeping everything around
[16:32:06] <anilg> I'm not sure how that is.. even if I got the entire source tree.. they still would be 12-15Gb..
[16:32:15] <anilg> and I dont have anything apart from sources there
[16:32:18] <rootard> not when they are unpacked
[16:32:28] <rootard> and there is a lot of duplication
[16:32:37] <anilg> the point is I dont have all the sources.. just some I worked on
[16:32:53] <rootard> since there is a tar.gz, a diff, and then an unpacked tree with binaries and then the .debs that ar built from all of it
[16:33:00] <anilg> I'll take a look and delete it. I just might keeps some of whats necessary.. and delete the entire dataset
[16:33:14] <rootard> thanks
[16:34:10] <mlo> no, not at runtime. Its a -D define while building, or patch to sys_defs.h
[16:34:36] <rootard> Then I vote for performance
[16:35:02] <anilg> rootard: cant devzone_enter..
[16:35:05] <anilg> [Connected to zone 'lanai' pts/14]
[16:35:05] <anilg> No utmpx entry. You must exec "login" from the lowest level "shell".
[16:35:06] <rootard> but that's just me voting :)
[16:35:20] <anilg> I am logging in from the lowest shell
[16:35:22] <anilg> ..
[16:35:24] <rootard> anilg: become root, zlogin -S zone_name
[16:35:47] <rootard> Since there is no free space, you can't login :)
[16:35:50] <mlo> so, on solaris, /devices/ and the symlinks in /dev are done dynamically, right ? can we just put those inodes in the .deb ?
[16:36:46] <rootard> They are generated dynamically, I don't know the implications of putting the device into the .deb
[16:37:44] <rootard> I do know that copying the dev from an installation script will fail in some circumstances where /dev/poll doesn't exist (for example a default zone)
[16:37:58] <mlo> As for the nscd thing, I have just reconfirmed: svcadm disable name-service-cache ; reboot --> Postfix comes up svcadm enable name-service-cache ; reboot -> postfix does not come up but complains about uids gids not matching stuff.
[16:38:13] <mlo> Note that this somewhat blows my theory out of the water that its just late in reading in the group file. Something else must be happening.
[16:39:08] <rootard> Maybe nscd can't be accessed in the chroot?
[16:39:13] <rootard> maybe a door is missing?
[16:39:54] <anilg> no free space to delete :-\
[16:40:07] <rootard> rm doesn't work?
[16:40:10] <anilg> # rm nexenta-on_5.11.104-1.tar.gz
[16:40:11] <anilg> rm: nexenta-on_5.11.104-1.tar.gz not removed: No space left on device
[16:40:11] <anilg> # echo $?
[16:40:11] <anilg> 2
[16:40:19] <rootard> oh, bcause of your snapshots
[16:40:23] <rootard> *because
[16:40:26] <anilg> ok..
[16:41:10] <anilg> great.. now that i have some breating space.. I'll go on a rm-ing spree
[16:41:21] <rootard> whee!
[16:44:00] *** kim0 has quit IRC
[16:47:52] <rootard> I'm glad my build finished before space was gone... it takes several hours to build python
[16:47:55] <rootard> BUT!
[16:47:56] <rootard> # file /usr/bin/python2.5
[16:47:57] <rootard> /usr/bin/python2.5: ELF 64-bit LSB executable, AMD x86-64, version 1 (SYSV), dynamically linked (uses shared libs), stripped
[16:59:54] <anilg> woo.. one particular directory was 80G.. it was filled with myspell-* packages
[17:00:13] <anilg> around 100G should be available now
[17:00:25] <mlo> rootard: This may be, but it happens on the master process already I would say. And it happens if i disable chroot on all services in master.cf
[17:01:26] <mlo> oh and btw, could you trigger a rebuild of gdb ? it seems it depends on libreadline5 5.1, when now only 5.2 is there. I wonder why the automatic dependencies dont catch that...
[17:02:35] <rootard> it needs to be rebuilt by hand. I'll add it to my list of things to do
[17:04:24] *** PatrickS has joined ##nexenta
[17:08:21] <rgl> postfix does not work in nexenta?
[17:08:43] <rgl> humm maybe I misunderstood the chat about postfix that just happened *G*
[17:13:38] <mlo> rgl: the standard postfix package is broken, at least for me. If yours works I would be interested in hearing about it
[17:14:42] *** NCommander has quit IRC
[17:16:16] <rgl> I did not try.
[17:16:36] * rgl is waiting for NCP 2 to retry nexenta :/
[17:23:42] <anilg> postfix doesnt install right because /usr/bin/rmail is also provided by a solaris package..
[17:24:23] <mlo> anilg: Yes, and somewhere there was already the magic line to add the dpkg diversion. Thats a non-issue.
[17:24:47] <mlo> (tho yes, that solaris package needs to be fixed)
[17:25:08] <mlo> the problem is that even once the thing is installed, it doesnt work.
[17:25:13] <mlo> For various reasons.
[17:25:22] <anilg> its sunwcsu.. thats core solaris.. not sure if mail should be core
[17:25:34] <anilg> hmm.. what are the issues you faced?
[17:26:08] <mlo> it doesnt link berkleydb (libdb) hence hash: maps dont work
[17:26:36] <mlo> The path for the modules is - on ubuntu and debian - /usr/lib/postfix, the internal paths however fetch the solaris default /usr/libexec/postfix
[17:27:16] <mlo> The configuration is for some of the daemons running chrooted, but the chroot is not set up, it lacks /dev/poll at the bare minimum
[17:27:39] <mlo> and - if nscd is running - it wont even start because it thinks postdrop has the same gid as nobody
[17:28:44] <mlo> there is probably minor details in that the ubuntu version has support for sdbm etc, which the default package also doesnt (because sys_deps.h is patched in the linux branch, and not in the solaris one)
[17:30:01] <anilg> I guess its better to use sun's postfix already provided in the base install
[17:41:52] *** kim0 has joined ##nexenta
[17:43:03] *** mib_2s8ka9 has joined ##nexenta
[17:43:55] <mib_2s8ka9> does the release of NCP2 on the topic mean that it goes from unstable to stable?
[17:44:27] <mlo> wait, sun does have a postfix in the base install ? On OpenSolaris ? More specifically, on Nexenta ?
[17:44:38] <mlo> I thought they were all sendmail ?
[17:45:46] <anilg> my mistake.. it's only sendmail
[17:46:00] <anilg> mib_2s8ka9: yes, with the final release
[17:46:21] <anilg> thats about 3 weeks away
[17:46:35] <mib_2s8ka9> has somebody here installed xfce4 or gnomecore succesfully with beta2?
[17:47:55] <anilg> yes.. but something seems broke with the latest bits. https://bugs.launchpad.net/bugs/336221
[17:49:22] <mib_2s8ka9> anilg: yes that was me :)
[17:50:12] <anilg> ow.. i know viridari recently setup xfce.. so you can hang around for him.
[17:50:30] <anilg> i dont have a box to try on, but I'll give this a shot in vmware a bit later
[17:51:48] <mib_2s8ka9> ok
[17:53:59] <mlo> did someone ever try to port fakeroot ?
[17:54:19] * anilg is building libgamin0.. hopefully that will resolve xfce
[17:57:24] <mib_2s8ka9> liggamin0 seems to be the underline cause for gnome-core too
[17:57:33] <mib_2s8ka9> libgamin0
[18:02:19] <anilg> mlo: i recollect giving it a shot
[18:02:32] <anilg> there are 2 versions in there tcp and sysv.. I got one to build.
[18:02:42] <anilg> the fix for the other would break the first
[18:06:10] <wurlitzer> t/opic
[18:06:17] <wurlitzer> oops
[18:06:47] <anilg> wurlitzer: you'd need to be an op to change the topic here, I think
[18:07:03] <wurlitzer> I know, I'm just checking the topic
[18:07:16] <wurlitzer> what is currently says
[18:18:00] *** mib_i3k7ov has joined ##nexenta
[18:25:48] <anilg> rootard: there..
[18:26:02] <rootard> anilg: ?
[18:26:19] <anilg> apt-cache search libgamin0 on nexenta.org
[18:26:21] <anilg> root@nexenta:/tank/ftproot/incoming/hardy# apt-cache search libgamin0
[18:26:22] <anilg> libgamin0 - Client library for the gamin file and directory monitoring system
[18:26:31] <anilg> on lanai after apt-get update
[18:26:38] <rootard> yes, it needs to be rebuilt
[18:26:43] <rootard> or brought back from the dead
[18:26:54] <anilg> # apt-cache search libgamin0
[18:26:55] <anilg> #
[18:27:01] <anilg> what happened?
[18:27:16] <anilg> how can i see it on one machine and not on the other?
[18:27:39] <anilg> same with gnome-terminal
[18:27:43] <rootard> apt-get update on nexenta.org and it should match
[18:28:04] <anilg> how did it go away in the first place.. newer package came in upstream?
[18:28:44] * anilg just built libgamin0, and is about to redput it
[18:31:52] *** anilg1 has joined ##nexenta
[18:42:03] <rootard> anilg: make sure you accept the fix into hardy-unstable _and_ hardy-testing
[18:50:07] *** anilg has quit IRC
[19:03:55] <anilg1> rootard: is there a particular argument to pass to redput2? I just used redput2 -p gamin
[19:04:13] <anilg1> it should finish in a few minutes
[19:05:30] *** anilg1 is now known as anilg
[19:08:09] <mlo> Uh, is nexenta completely auto-built in some way ?  Who can make changes there ? fetchmail is not in the repo. Building it from scratch fails, due to some kerberos check. It however goes through with PATH including /usr/sun/bin (krb5-config is there).
[19:08:19] <mlo> So what needs to be done that this (and similar packages) are being autobuilt ?
[19:10:00] <anilg> not completely.. not a lot of non-main packages are built using the autobuilder
[19:10:11] <anilg> http://builder.tajinc.org/
[19:10:37] <anilg> these are only packages that build cleanly.. if an upstream source build fails..
[19:10:47] <anilg> we need to fix the issue, or build it manually and upload
[19:11:54] <anilg> in the case of fetchmail.. if it fails with prestine sources, but builds fine with certain changes.. these packages and modified sources are uploaded manually
[19:12:29] <anilg> so the next person who wants to build it from source can simply pull in these modified sources and build away
[19:14:44] <mlo> well, actually, it builds fine from apt-get source fetchmail, but only if /usr/sun/bin is in $PATH. So how is this to be fixed ?
[19:16:12] <anilg> you add /usr/sun/bin to the path in debian/rules
[19:18:36] <rootard> is kerberos installed?
[19:19:57] <rootard> There should be a gnu equiv for krb5-config
[19:20:29] <anilg> heimday
[19:20:33] <anilg> heimdal*
[19:21:17] <rootard> We actually should build against krb5 if possible to keep thing consistent
[19:21:35] <rootard> php5 had to be rebuilt against krb5
[19:22:35] <rootard> Also, Sun's krb is MIT Krb5 as well
[19:22:40] <mlo> I think krb5 is whats installed on ubuntu too. I know nothing that uses heimdal.. booting ubuntu to check whats where there.
[19:23:03] <mlo> Suns KRB is mit krb5, and krb5-config is in /usr/sun/bin and suffices to build fetchmail; were it in the path.
[19:24:07] <rootard> in general I try to build GNU software against GNU packages where possible.
[19:24:50] <mlo> Ah, heres the problem.
[19:25:15] <mlo> nexenta libkrb5-dev has a "libkrb5-config" which probably should be krb5-config. Ubuntu has krb5-config, and no libkrb5-config
[19:26:47] * rootard recalls changing that some time ago
[19:27:12] <rootard> ie: so Nexenta had krb5-config and not libkrb5-config
[19:27:38] * rootard wonders where the change went/came back from
[19:27:59] <rootard> gdb rebuilt and uploaded
[19:28:32] *** mib_i3k7ov has quit IRC
[19:28:45] <mlo> no. Nexenta has libkrb5-config, which should be named krb5-config
[19:29:07] <rootard> mlo: we are saying the same thing :)
[19:30:14] <rootard> anilg: passing -d hardy-testing  to redput2 _should_ do the trick
[19:30:20] <rootard> I haven't tried it yet
[19:30:55] <anilg> now that i've run redput2.. the package is deleted..
[19:30:58] *** dafinga has quit IRC
[19:31:08] <anilg> shall I reupload and run with -d?
[19:31:17] <rootard> yes
[19:31:40] <rootard> I think there is a switch to keep the package upload around
[19:33:38] <anilg> BTW, why are we also keeping testing in mind now?
[19:34:00] <rootard> so we can accept packages into unstable if we are unsure about them
[19:34:05] <rootard> only fixes should go into testing
[19:34:20] <rootard> and things that are not working should be ripped out of testing
[19:36:03] <rootard> At release time, testing will turn into stable. I don't want to call it stable though... because it isn't :)
[19:37:00] <rootard> no, it's not
[19:38:17] <rootard> ie, the problem is not with redput2, it is with snapshots
[19:39:51] <anilg> the command finished instantaneously.. so i'm not sure if it was correctly uploaded
[19:39:59] <anilg> exit code was 0
[19:40:05] <rootard> probably not
[19:40:10] <mlo> So how are the changed rules files etc. kept ? Is there some svn repo for it ? Or is it simply documented in the changelog for the next time ?
[19:41:28] <anilg> "simply documented in the changelog for the next time"..
[19:41:30] <anilg> or in the diff
[19:41:41] <anilg> which serves as a reference for next time
[19:41:46] <rootard> mlo: I think we need a better system
[19:42:55] <rootard> the best solution I have seen proposed seems to be based mercurial patch queues
[19:43:01] <rootard> *based around
[19:44:11] <mlo> OK, so if I were to help, what would I do ? According tho BuildingPackages I can just create a gpg key, build packages, and dput them ? I do not need someone to sign my key etc ?
[19:44:45] <rootard> you need to give us your pub key so that package uploads are verified
[19:45:07] <rootard> but yes, that is the way things are currently done
[19:47:05] <mlo> Is that the same when using the available autobuilder ? I take it in that case the gpg key would have to be stored on that machine. In this case I certainly wouldnt submit that to the pgp keyserver, nor would I put my personal key there
[19:47:47] *** dafinga has joined ##nexenta
[19:47:51] <rootard> anilg: am I safe to run redput2?
[19:48:22] *** fserve has joined ##nexenta
[19:53:33] <mlo> anilg: I tried looking at the libkrb5-dev. I cannot rebuild it due to a missing build-dependency to ss-dev   how did you build that package ??
[19:54:15] <rootard> mlo: the ab-signs with its own key
[19:56:03] <anilg> rootard: yes.. i did not try with hardy-testing again
[19:59:25] <mlo> ok. So who gives me access to the build machines ?
[20:00:55] <anilg> I can :)
[20:01:13] <anilg> our machine is at gnusolaris.org
[20:01:36] <anilg> you get a user account, and once logged in you can create a devzone wtih full root access to do as you please
[20:01:47] <rootard> *within reason*
[20:01:56] <anilg> yes :)
[20:02:12] * anilg just likes harping about how users can do root level stuff with devzones
[20:02:26] <mlo> "heres these hot GNU warez...."   ;)
[20:02:46] <anilg> what login name do i set you up with?
[20:02:57] <mlo> mlo
[20:03:29] <mlo> The thing is that the zones werent all tooo shiny. Linux has had them for years as vservers ...
[20:03:54] <rootard> mlo: yes, but zones are vservers on steriods
[20:04:20] <mlo> Other than brandz zones, what can you do with a zone you cant on vserver ?
[20:04:30] <rootard> vservers (while neat/fun) are a hack compared to zones
[20:04:37] <rootard> yes
[20:04:44] <rootard> all kinds of resource allocation
[20:05:10] <rootard> you can setup a nic for a zone
[20:05:42] <rootard> bringing other devices into a zone can be done via zonecfg (something else that vservers lack)
[20:06:15] <anilg> think of the things crossbow's network virtualization with zones can do :)
[20:06:20] <rootard> then of course you have zfs fun (as you will soon find out)
[20:06:26] <anilg> can do with zones*
[20:06:50] <rootard> once you get your devzone, run "zfs list" and you will notice you have your very own dataset to use
[20:07:39] <rootard> you can create datasets inderneath it "zfs create -o compression=gzip-9 tank/users/mlo/archive;zfs set mountpoint=/archive tank/users/mlo/archive"
[20:08:42] <rootard> ...
[20:09:32] <anilg> I'm helping setup his account and devzone
[20:09:34] <rootard> can you run "reboot" in a vserver and reboot the vserver?
[20:09:55] <mlo> zfs fun is the reason why I am still trying to get nexenta working. Otherwise I would have just sticked to linux, and I am contemplating testing freebsd again
[20:10:42] <mlo> reboot in vserver works, and you can set up virtual nics, with their own ip routing, and firewall rules too iirc
[20:16:00] *** tsukasa` has quit IRC
[20:16:37] <fserve> mlo
[20:16:44] <fserve> nexenta isn't only zfs
[20:16:50] <fserve> apt-clone is very fun too : )
[20:17:02] <fserve> and i liked debtorrent
[20:17:44] *** tsukasa has joined ##nexenta
[20:30:40] <anilg> night all.
[20:30:53] *** anilg is now known as anilg_Zzz
[20:32:09] <mlo> so what is the idea about versioning? Use the same version as ubuntu, extending the changelog of the previous (ubuntu) maintainer ?
[20:32:48] <anilg_Zzz> mlo: ubuntuX changes to nexentaX+1
[20:33:15] <anilg_Zzz> any other ending.. append nexenta1 to it
[20:33:22] <anilg_Zzz> dch -i usually does the right thing
[20:33:41] *** mib_uq4jvv has joined ##nexenta
[20:34:16] <anilg_Zzz> ok.. really Zzz now
[20:39:58] <mlo> ok, so I am now at apt-upstream-tool. The package is marked "feisty" there, what should it be for NCP ?
[20:40:34] <rootard> hardy
[20:40:45] <rootard> NCP2 == hardy
[20:40:52] <rootard> NCP1 == elatte
[20:44:19] <mlo> ok, when creating the gpg key, do I do that on gnusolaris, or do I do that in the zone, and do I have to re-import the key in the zone every time ?
[20:45:06] <rootard> mlo: it's up to you.
[20:45:16] <rootard> its easiest if you keep it inside the zone
[20:45:20] <rootard> on the zfs dataset
[20:45:39] <rootard> (the zfs dataset persists between devzone_free/devzone_create
[20:45:40] <rootard> )
[20:50:02] *** mib_uq4jvv has quit IRC
[20:50:42] <mlo> I take it my key has to be my real email addess, not mlo@gnusolaris ?
[20:56:43] <rootard> it is helpful to use a real address...
[20:58:06] <mlo> Thats not the point. The point is mainly I will keep the nexenta key separate from my usual key. So I do have to use a separate address anyway.
[20:58:36] <mlo> OK, so how do I copy data from the gnusolaris login to my zone, respectively the /archive dataset ?
[20:59:07] <rootard> you can create a subkey
[20:59:12] <rootard> and then import that
[20:59:18] <rootard> or you can always sign the package remotely
[20:59:40] <rootard> scp should work
[20:59:51] <rootard> you can use /archive or create another dataset
[20:59:56] <mlo> scp to where ?
[20:59:58] <rootard> it's your env... use it as you see fit
[21:00:25] <rootard> (or maybe I'm not understanding the question...)
[21:02:01] <mlo> The question is: I am logged on as mlo@gnusolaris. I have some files there (namely the gpg key). How do I copy that into my dev zone ?
[21:04:24] <mlo> ah ok, I can pull it from the devzone via scp
[21:04:39] <rootard> enter the devzone (devzone_enter); figure out where you want to place the files (cd /root); and then use scp to copy them (scp -r mlo at gnusolaris dot org:.gnupg ./)
[21:07:21] <mlo> thats what i ultimately did.
[21:16:09] <mlo> Is: Maintainer: Ubuntu Core Developers <ubuntu-devel-discuss at lists dot ubuntu.com> still correct ?
[21:16:43] <rootard> yes
[21:18:04] *** rgl has quit IRC
[21:18:58] <mlo> OK. It says successfully uploaded package. Now I watch the autobuilder ?
[21:27:15] <mlo> So whats the word regarding daemons and smf ? Is the usual init script the preferred way, or is it preferred to write smf classes ?
[21:29:52] *** anilg has joined ##nexenta
[21:31:16] <rootard> mlo: if you loaded a source-only package then the autobuilder would build it. Since you uploaded a source+binary the autobuilder has nothing to do.
[21:31:52] <mlo> I am not exactly sure what I uploaded, I just followed the instructions :)
[21:33:47] *** anilg1 has joined ##nexenta
[21:48:11] *** anilg_Zzz has quit IRC
[21:50:47] *** anilg has quit IRC
[22:24:18] *** mlo has quit IRC
[22:29:42] *** mib_r9wg6e has joined ##nexenta
[22:30:33] *** mib_r9wg6e has quit IRC
[22:41:29] *** mib_2s8ka9 has quit IRC
[23:12:11] *** Corwin has joined ##nexenta
[23:43:37] *** lesterc has joined ##nexenta

   March 1, 2009  
< | 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 | >