=== khermans [n=khermans@lambda.sbs.umass.edu] has joined #ubuntu-server === hunger [n=hunger@p54A62815.dip0.t-ipconnect.de] has joined #ubuntu-server === etcp [i=foobar@home.etcp.co.uk] has joined #ubuntu-server === zenrox [n=zenrox@71.115.198.118] has joined #ubuntu-server === zenrox [n=zenrox@71.115.198.118] has joined #ubuntu-server === hunger [n=hunger@p54A6394F.dip0.t-ipconnect.de] has joined #ubuntu-server === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === lbm [n=lbm@cpe.atm4-0-1301006.0x50a0824e.vgnxx6.customer.tele.dk] has joined #ubuntu-server === fabbione [n=fabbione@port49.ds1-van.adsl.cybercity.dk] has joined #ubuntu-server [09:12] This 2.6.15 kernel is strange! [09:13] Yesterday night the -6-686 did not boot before I went to bed. When I woke up today it did. [09:17] The -6-686 works way better than the -5-686 though === zenrox [n=zenrox@71.115.198.118] has joined #ubuntu-server === spike [n=spike@unaffiliated/spike] has joined #ubuntu-server [11:13] 'morning [11:15] morning guys === ..[topic/#ubuntu-server:fabbione] : ubuntu-server discussions and support | for general support see #ubuntu or https://wiki.ubuntu.com/InternetRelayChat for other channels | Download Ubuntu Server Edition at http://releases.ubuntu.com/ubuntu-server/5.10/ | Ubuntu Server dapper specs: https://launchpad.net/people/ubuntu-server/+specs | Ubuntu Server forums at http://ubuntuforums.org/forumdisplay.php?f=45 [11:16] troy: please leave a pointer to the specs [11:17] given that's what we are going to play for dapper === spike checks the specs [11:25] fabbione: is a ML available or it's still a work in progress? [11:25] spike: i did ask again yesterday to the admins [11:25] but other than that i have no news [11:29] fabbione: woooo, finally the MD5 checker, god bless you man :) [11:30] spike: i have the core already implemented [11:31] i hope to finish it sometimes next week [11:32] A dtrace port or use/investigate/integrate/package system tap (apparently a dtrace-alike for Linux) <------ that doesnt work afaik [11:32] systemtap I mean. it's under dev, but far from being enterprise ready [11:32] last time I checked at least, might check again, I was after it too [11:33] (playing with dtrace with solaris10/nexenta) [11:33] spike: take into account that some stuff there is really in "wishlist" state [11:33] not all the spec will and/or can be implemented for dapper [11:33] but since we had the ideas, we wrote all of them down [11:33] anyway... lunch :) [11:33] i am hungry [11:34] fabbione: about /etc under RCS... would you think that throwing in cfengine would be too much of a trouble? [11:34] ok, talk to u later === spike [n=spike@unaffiliated/spike] has joined #ubuntu-server === sebest_ [n=sebest@79.245.101-84.rev.gaoland.net] has joined #ubuntu-server [01:26] hello [01:27] hi there [01:28] to introduce myself, i'm one of the developpers of avahi, and i wrote the "corporate users" part of this page https://wiki.ubuntu.com/ZeroConfSpec/Misc [01:28] and i was interested in a ubuntu for server with the same spirit as its desktop counterpart === Yagisan [n=jamie@60-240-201-251.tpgi.com.au] has joined #ubuntu-server [01:30] sebest_: cool, I was lookin at that few days ago, really nice stuff [01:31] spike, yes will make ubuntu even easier to use :) [01:31] if we can get a server that works nicely with the dekstop, it would be a really great thing for small business === Valandil [n=chrys@dslb-084-056-107-052.pools.arcor-ip.net] has joined #ubuntu-server [01:33] sebest_: I agree. not my personal first interest, but indeed nice [01:33] 'lo Valandil [01:38] fabbione, the RCS idea for the servers is great, i thought about it previously using svn (but bzr is equivalent) but why does it need the crontab script? [01:38] with subversion you can have action on commit [01:38] sebest_: it's not a problem of action on commit [01:38] eg: call a script that check which file as been altered and restart the coressponding service [01:38] no no.. you are on the wrong track [01:39] the crontab is to make the commit transparent to the admin [01:39] let say you start editing files [01:39] and you forget to commit [01:39] the crontab will commit for you [01:39] with a timestamped commit [01:39] but no service restart [01:39] that's up to the admin [01:39] but the idea is good.. it could be a config option === spike proposed a stronger solution cfengine like [01:40] even if the hook would be pretty complex [01:40] and that would help with stuff like https://wiki.ubuntu.com/NetworkWideUpdates [01:40] ah ok [01:40] spike: i don't know cfengine and how complex it is [01:40] so it would be, imho, a better approach for further development [01:40] spike: none of the admins did ever mention it [01:40] during the spec writing.. that's it [01:40] fabbione: I guess I can contribute on that.. [01:40] spike: the specs are "closed" for dapper [01:41] you are welcome to add comments and ideas at the bottom of the wiki [01:41] fabbione: well, imho that's the standard for large network management, as accepted by usenix/sage [01:41] another nice thing with RCS, is versionning (rollback) and log message to explain the chaange [01:41] spike: please add it at the end [01:41] spike: so there is a record [01:41] and we will not forget [01:41] they've also got dpkg hoos, to install pkgs and such [01:41] hooks* [01:41] ok [01:42] sebest_: the idea behind RCS is exactly to give admins the possibility to rollback easily [01:43] each admin will have an etc in their home, so they can edit there local copy, right? [01:43] they don't need to sudo or su to edit config file? [01:43] fabbione: to this page? https://wiki.ubuntu.com/ServerCandy [01:43] in theory [01:43] you could also push changes between differnt machines [01:43] like you do a change on a machine [01:43] sshcluster to do a bzr cherrypick that changeset [01:43] spike: yes [01:43] so that for example your apache config is updated everywhere on the production servers from the staging/testing machine [01:43] would it be possible to give access to some part of /etc/ selectively [01:44] eg not allow someone to alter apache config, and another one postfix [01:44] sebest_: that's a bzr feature we are missing [01:44] nested trees... [01:44] basically bzr archives inside bzr archives [01:44] you achive that possibility once there are nested trees and you can block access via normal unix permissions [01:45] hum, is it a planned feature? [01:45] yes [01:45] but there is no ETA on it yet [01:45] mmmh, doesnt svn gives you ability to restrict access to subdir, does it? I seem to recall such thing [01:45] plus we need to consider that we need to use a stable version of bzr === spike never used bzr [01:46] spike: not that i know of [01:46] fabbione: I'll check that out [01:46] and in any case svn has the problems that it requires a running server [01:46] bzr doesn't [01:46] a running server? [01:46] well you need to setup the server side of svn [01:46] you mean for remote access? [01:46] even the local db [01:46] no [01:46] svn admin create toto [01:47] svn co file://path/toto [01:47] sebest_: but than you lose other features [01:47] ah maybe, i don't know [01:47] like the possibility to cherry pick from server to another [01:47] i know that i use it without server === spike avoid commenting 'till he learned more about bzr [01:47] i use it over ssh [01:47] cherry pick? [01:47] i admin that my svn experience is client side only === spike uses svn server and likes it a lot [01:48] spike: yes.. never done cherry picking? === sebest_ too [01:48] ssh has restrictions, principally no virtual users [01:49] fabbione: oh, ok, just googled the term. no, I havent, never had the need [01:49] i know that svn can have lock problem when using db format [01:50] it's better to use ffs [01:50] ok [01:50] but we are not going on svn.. i can tell you that from now [01:50] i know nearly nothing about bzr yet [01:50] sebest_: fsfs solves that [01:50] yep [01:50] plus you need to keep in mind one thing [01:51] this miniRCS thingy [01:51] but i'm ok with any RCS system as long as it fullfill our need :) [01:51] is meant to be used only for "recovery" [01:51] it's not something you are going to use regularly to do your development [01:51] think about it like a daily backup [01:51] except a bit more powerful :) [01:52] the advantages that i saw what the loggin stuff (each admin, can write a message about why he made a change) [01:52] well,once u've it up it should be up to u to use it whatever u want it for, no? I don't see why designing something like an RCS system restricted to "recover" [01:52] so it's easier to trace wrong change [01:52] sebest_: you don't lose that [01:52] and also easier to have many admins working on the same server [01:52] sebest_: the cron is there only if changes are done and you forget to commit [01:52] if there are no changes, there is no commit [01:52] fabbione, yes i got it :) [01:52] fabbione: btw, why bzr? [01:53] spike: it's distributed, doesn't require a central db/server and we support it directly :) [01:53] and, do you think the infrastructure will be modularized enough so one can replace bzr with something else? [01:53] possibly yes [01:53] i didn't get to write that code yet [01:54] because, now that u mentioned changes and multiple admin, I'm working on stuff to wrap things like that in a ticketing system, trac like, because that's what you relly want to large enterprise imho [01:54] and svn is nicely integrated with trac, so to me, bzr is a problem from that POV [01:54] spike: svn is not distributed and that gives me tons of other problems [01:55] spike: i can write it modular.. you can write the svn plugins [01:55] but i am not going to invest time on RCS != bzr [01:55] fabbione: whatever, guess the spec should just allow switching. I never had a prob with it, I guess because mine and yours requirements are different [01:56] fabbione: that's perfect for me, I'll invest time to support svn, I'd just like the specs to allow me to do so [01:56] spike: as it is they don't.. but that's a code design issue [01:56] spike: there is no need for the specs to be modified [01:56] i can just do that at code level [01:56] and allow a certain level of modularization [01:57] fabbione: ok, "specs" was just a name for it,whereever you wanna do it, I'm fine with it [01:57] spike: so am i :) [01:58] fabbione: do you mind an OT question? are you familiar with telecom interbusiness router issues when it comes to additional subnets? [01:58] spike: telecom italia? [01:58] imho they do something really nasty here, but they have exclusive access to the router, and I'm not sure how to convince them they have to reconfigure it [01:58] fabbione: yes [01:59] spike: no sorry [01:59] i don't live in italy [01:59] and i did stop working for seabone a loooooong time ago [01:59] good for you :), actually not either I will anymore in a month :D [02:06] fabbione, what kind of help do you need for the server related stuff on ubuntu? [02:08] personnality i've interest in the SBS stuff [02:09] this may be a particular use case of the server project [02:10] sebest_: it depends what you can do [02:10] i go eating, we'll discuss this later [02:11] i might not be around later [02:11] but ok [02:11] enjoy your lunch [02:11] i can do a lot things ;) (skills in admin, packaging and programming C/php/python) [02:11] php isnt really programming :P === spike <--- evil [02:11] :D [02:11] that's why i added pythoon ;) [02:12] brb [02:19] hi all :-) === mgalvin [n=mgalvin@ubuntu/member/mgalvin] has joined #ubuntu-server === Yagisan [n=jamie@60-240-201-251.tpgi.com.au] has joined #ubuntu-server [03:28] re === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === henriquemaia [n=henrique@cb-217-129-170-53.netvisao.pt] has joined #ubuntu-server === hunger [n=hunger@p54A61C86.dip0.t-ipconnect.de] has joined #ubuntu-server [04:31] re === Valandil [n=chrys@dslb-084-056-107-052.pools.arcor-ip.net] has joined #ubuntu-server