November 4, 2013 
< | 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 | >
November 4, 2013 [00:05:39] *** dgo has quit IRC[00:14:39] *** hlangeveld has quit IRC[00:31:28] *** szaydel has joined #omnios[00:39:52] *** cantstanya has joined #omnios[00:41:00] *** cantstanya has quit IRC[00:47:45] *** cantstanya has joined #omnios[00:48:14] *** cjones_ has quit IRC[00:54:36] *** Clory has joined #omnios[01:03:43] *** cjones_ has joined #omnios[01:13:13] *** szaydel has quit IRC[01:15:35] *** szaydel has joined #omnios[01:35:01] *** szaydel has quit IRC[01:49:43] *** szaydel has joined #omnios[01:53:37] <kimc> Now have Apache2.4.6 dovecot 2.2.6 postfix 2.10.2 with associated smf's running fine[01:58:00] *** szaydel has quit IRC[02:14:57] *** kimc has quit IRC[03:05:26] *** cantstanya has quit IRC[03:18:41] *** cantstanya has joined #omnios[04:10:00] *** iAmTheDave has quit IRC[04:40:52] *** djun-bb has joined #omnios[04:53:45] *** nikolam has quit IRC[05:00:13] *** ball has joined #omnios[05:12:16] *** djun-bb has quit IRC[07:08:46] *** cjones_ has quit IRC[07:15:19] *** dgo has joined #omnios[07:15:53] *** ball has quit IRC[07:16:37] *** nefilim has quit IRC[07:58:27] *** slx86 has joined #omnios[08:01:42] *** slx86 has quit IRC[08:11:08] *** slx86 has joined #omnios[08:36:27] *** hlangeveld has joined #omnios[08:46:18] *** vrou has joined #omnios[09:26:26] *** Clory has quit IRC[09:26:51] *** hydan has quit IRC[09:31:30] *** Clory has joined #omnios[09:56:00] *** bens1 has joined #omnios[10:26:14] *** Tom_H has joined #omnios[10:26:15] *** Tom_H has joined #omnios[10:50:24] *** slx86 has quit IRC[10:50:48] *** slx86 has joined #omnios[11:22:10] *** KermitTheFragger has joined #omnios[11:32:07] *** nikolam has joined #omnios[12:11:04] *** datadigger has quit IRC[12:11:12] *** datadigg1r has joined #omnios[12:27:26] *** Tom_H has quit IRC[12:28:02] *** Tom_H has joined #omnios[12:33:48] *** khushildep has joined #omnios[12:42:54] *** digidigger has quit IRC[12:43:19] *** digidigger has joined #omnios[12:44:13] *** khushildep_ has joined #omnios[12:46:23] *** khushildep has quit IRC[12:46:23] *** khushildep_ is now known as khushildep[13:25:17] *** vrou has quit IRC[14:00:11] *** cjones_ has joined #omnios[14:05:03] *** szaydel has joined #omnios[14:35:18] *** postwait has joined #omnios[14:35:20] *** ChanServ sets mode: +o postwait[14:40:56] <lotheac> are omniti-perl build scripts not public or do I just suck at finding them?[14:41:15] *** pygi has joined #omnios[14:46:25] <postwait> not sure where they are.[14:46:34] <postwait> they may not be public (for no good reason)[14:49:18] <postwait> lotheac: in fact, I don't know where it is.[14:49:24] <lotheac> heh :)[14:49:36] <lotheac> just kinda interested[14:49:54] <lotheac> also I'm seeing a discrepancy between PERL_MAKEFILE_OPTS being defined in config.sh and MAKEFILE_OPTS being used in functions.sh...[14:50:19] <lotheac> and neither one being used anywhere else[14:50:40] *** toebu has joined #omnios[15:02:06] *** nikolam has quit IRC[15:09:47] *** mihai_omniosuser has joined #omnios[15:09:47] *** postwait has quit IRC[15:19:51] *** khushildep_ has joined #omnios[15:22:36] *** khushildep has quit IRC[15:22:37] *** khushildep_ is now known as khushildep[15:28:08] <lotheac> postwait, also, thoughts on that pkgrepo permissions thing?[15:51:09] *** neophenix has joined #omnios[15:51:40] *** slx86 has quit IRC[16:02:25] *** Tom_H__ has joined #omnios[16:04:12] *** Tom_H has quit IRC[16:05:31] *** daleg has joined #omnios[16:09:27] *** porkotron has quit IRC[16:10:45] *** daleg_ has joined #omnios[16:12:06] *** daleg has quit IRC[16:12:06] *** daleg_ is now known as daleg[16:12:34] *** theason has quit IRC[16:23:20] *** toebu has quit IRC[16:32:29] *** theason has joined #omnios[16:37:28] *** toebu has joined #omnios[16:41:06] *** szaydel has quit IRC[17:01:15] *** toebu has quit IRC[17:02:05] *** szaydel has joined #omnios[17:04:21] *** toebu has joined #omnios[17:20:58] *** Tom_H__ has quit IRC[17:21:33] *** Tom_H has joined #omnios[17:24:38] *** cantstanya has quit IRC[17:34:23] *** pygi has quit IRC[17:35:29] *** postwait has joined #omnios[17:35:31] *** ChanServ sets mode: +o postwait[17:37:20] *** cantstanya has joined #omnios[17:41:03] *** cjones_ has quit IRC[17:54:44] *** pygi has joined #omnios[17:56:09] *** desai has joined #omnios[17:56:47] *** pygi has quit IRC[18:00:33] *** Hydan has joined #omnios[18:00:57] *** Hydan is now known as Guest69286[18:01:10] *** Guest69286 is now known as hydan`[18:01:37] *** hydan` is now known as hydan[18:04:07] *** desai has quit IRC[18:06:00] *** cjones_ has joined #omnios[18:07:46] *** desai has joined #omnios[18:08:13] *** nikolam has joined #omnios[18:15:39] <lotheac> hrm, looks like you weren't actually here when I asked... < lotheac> postwait, also, thoughts on that pkgrepo permissions thing?[18:16:05] <postwait> lotheac: two thoughts.[18:16:17] <postwait> (1) I think removing all checks doesn't make much sense.[18:16:31] <postwait> (2) you can work around your problem today with extended ACLs in ZFS.[18:16:45] <postwait> So.. I think it might make sense to relax it to 664...[18:16:53] *** iAmTheDave has joined #omnios[18:16:56] <lotheac> 1) I conditionally agree; that is, if I knew why they are there :P[18:17:22] <lotheac> 2) really? I couldn't find any aclmode that didn't *also* affect the file mode when I set the necessary ACLs[18:18:02] <postwait> It may interfere with new file creation.[18:18:22] <postwait> but you should be able to have users create 644 files in a directory they don't own using those ACLs[18:18:37] <postwait> I think it is unsafe to have files world writable.[18:18:42] <postwait> it really makes no sense.[18:19:11] <lotheac> yes, I agree, but I don't think it's the job of the packaging tool to manage the permissions[18:19:36] <lotheac> oh, hmm, maybe I could use acls with added perms for a certain group instead of using group@ to work around it[18:19:52] <lotheac> but if I use group@ it will affect the modes[18:20:44] <lotheac> I'll test tomorrow, thanks[18:24:50] *** joltman has joined #omnios[18:27:09] <postwait> keep in mind... it isn't managing.[18:27:15] <postwait> it's protecting.[18:27:25] <postwait> managing (in my mind) would be fixing the permissions.[18:27:58] <postwait> sshd's job isnt' the manage permissions on config files... but it shouldn't run when it believe unsafe circumstances.[18:28:11] <postwait> that said.. 664 solves your problem, no?[18:36:19] *** desai has quit IRC[18:37:50] *** Tom_H has quit IRC[18:39:50] *** Clory has quit IRC[18:40:17] *** Clory has joined #omnios[19:20:35] <lotheac> postwait: but it *does* also fix the permissions if it's not in readonly mode (such as 'pkgrepo info')[19:20:41] <lotheac> well, "fix", it sets them to what it expects[19:20:50] <lotheac> and yes, 664 would fix it[19:20:52] <postwait> interesting.[19:20:59] <postwait> that is unexpected.[19:21:21] <postwait> I'll note your use-case doesn't match ours.[19:21:23] <lotheac> I didn't actually test if it does that though but there's comments to that effect in the source[19:21:32] <postwait> we just run a pkgrepo server and collectively publish to it.[19:21:43] <lotheac> yes, I know, but I don't want to run a http reverse proxy in front of it just to publish[19:21:49] <postwait> We don't[19:22:05] <lotheac> and there's even a limitation with http publishing not being able to handle >128MB files documented in pkgsend man page ;)[19:22:09] <postwait> only the ones are are publicly accessible[19:23:03] <lotheac> well, I want some form of access control and filesystem access is pretty much perfect for us because it's not intrusive (no separate authentication needed)[19:23:34] <lotheac> we do use http for downloading from it but that's completely different of course :)[19:23:56] <lotheac> might even consider it a security feature that the depotd cannot write to it[19:25:03] *** porkotron has joined #omnios[19:29:39] <lotheac> I will note that depotd+file system publishing sucks in other ways, like needing to restart depotd every time you do something[19:29:59] <lotheac> but I'd still prefer it over http[19:32:43] <lotheac> and oh, just changing the mode it expects to 664 is probably not a good idea because while group writable makes sense in our env, it might not everywhere, and seeing how it will "fix" the existing perms...[19:33:08] <lotheac> so not really sure what it makes sense for it to check.[19:36:40] *** KermitTheFragger has quit IRC[19:39:37] *** gmason has joined #omnios[19:40:19] <postwait> lotheac: I don't seem to have that issue.[19:40:37] <postwait> I publish to depotd over http remotely and hardly ever log in to administer/restart.[19:41:31] <lotheac> well, yes, I know that it works over http because then depotd will obviously notice it has changes :P[19:41:39] <lotheac> it's documented that it won't over filesystem publishes though[19:43:39] <lotheac> still, not much of an issue; pkg refreshes and updates and everything will still work without restarting depotd but the web interface won't know about new stuff if you don't restart it[19:43:43] *** esproul has joined #omnios[19:43:44] <lotheac> so it's only a minor issue[19:43:54] <lotheac> looks like I'm being redundant today[19:45:36] <postwait> yeah... filsystem and http don't mix.[19:45:49] <postwait> which is why I don't ever interact with depos over fs.[19:46:06] <postwait> though I occassionally publish to a local fs repo and pkgsend it to a remote one...[19:46:41] <lotheac> I could make the same argument for not interacting with them over http, but there's a bug in that scenario which prevents shared repo administration ;-)[19:50:08] <lotheac> anyways I can see the point in making sure files in the repo are not world writable, but then I can see the point in other environments that they are not group writable either. thing is it's pretty dumb to hardcode it[19:51:25] <lotheac> so if it's relaxed to allow group write there's less point for even having it (won't protect someone who has their files' group set to 'users' or something)[19:54:58] <lotheac> but then again if a specific acl can work around it then this is all pretty much a moot point[19:55:51] <lotheac> though in the presence of acls not presentable by file modes the permission sanity check doesn't have much point either :P[20:03:02] *** joseyluis has joined #omnios[20:04:40] *** khushildep has quit IRC[20:05:17] <joseyluis> I am trying to install the last stable release of omnios: omnios-b281e50. I have two disks of 4 Tbytes. The installation complains that it can't use more of 2 Tbytes. Do you know if has been solved in bloody release. I suppose that it is the same limitation of Solaris 10 with boot disks.[20:05:34] <joseyluis> Do you know if has been solved in bloody release or some workaround?[20:10:29] *** wez has joined #omnios[20:28:14] *** desai has joined #omnios[20:30:00] *** toebu has quit IRC[20:54:19] *** philips has quit IRC[20:56:35] *** philips has joined #omnios[21:00:12] *** pygi has joined #omnios[21:10:43] *** Tom_H has joined #omnios[21:25:44] *** pygi has quit IRC[21:35:24] *** wez is now known as wez|away[21:50:45] *** wez|away is now known as wez[22:01:15] *** iAmTheDave_ has joined #omnios[22:01:57] *** iAmTheDave has quit IRC[22:01:58] *** iAmTheDave_ is now known as iAmTheDave[22:10:35] *** klobucar has quit IRC[22:19:05] *** _jared_ has quit IRC[22:19:09] *** gregburd has quit IRC[22:19:14] *** cheeseplus has quit IRC[22:21:43] *** wez is now known as wez|away[22:22:51] *** thom has quit IRC[22:33:33] *** wez|away is now known as wez[23:03:15] *** gmason has quit IRC[23:04:01] *** gmason has joined #omnios[23:06:13] *** klobucar has joined #omnios[23:06:59] *** iAmTheDave has quit IRC[23:08:28] *** gmason has quit IRC[23:26:08] *** gmason has joined #omnios[23:26:30] *** desai has quit IRC[23:32:06] *** gmason has quit IRC[23:39:07] *** hydan has quit IRC[23:39:39] *** _jared_ has joined #omnios[23:42:27] *** gregburd has joined #omnios[23:43:04] *** cheeseplus has joined #omnios[23:53:38] *** joltman has quit IRC