[03:06:51] 07/10/2012 - 03:06:51 - User laner may have been modified in LDAP or locally, updating key in project(s): deployment-prep [06:10:10] 07/10/2012 - 06:10:10 - Creating a home directory for jesusaurus at /export/keys/jesusaurus [06:10:31] so, ecdsa is not a valid key type? [06:10:38] :( [06:11:11] 07/10/2012 - 06:11:11 - Updating keys for jesusaurus at /export/keys/jesusaurus [06:38:11] 07/10/2012 - 06:38:11 - Updating keys for jesusaurus at /export/keys/jesusaurus [10:49:07] A question: How and who can start a new Labs-Project and under which conditions? [11:06:16] A question: How and who can start a new Labs-Project and under which conditions? [11:08:58] There's a few people, fire a mail to the list and someone will sort it out [11:12:16] to the lists.wikimedia.org-List? [14:10:10] 07/10/2012 - 14:10:10 - Creating a home directory for jasonspriggs at /export/keys/jasonspriggs [14:11:11] 07/10/2012 - 14:11:11 - Updating keys for jasonspriggs at /export/keys/jasonspriggs [14:18:02] @andrewbogott: Hi [14:18:36] Jan_Luca: howdy! Are you still in need of labs privs? [14:19:45] Special:NovaKey seems to work now but I'm not in bastion-group [14:19:51] And, let's see... do you have a gerrit account already? If so, what's your username there? [14:19:59] (Special:NovaKey fixed a relogin) [14:20:08] Yes, name: jan [14:20:50] Did you need someone to create a project for you as well, or is that covered? [14:21:08] Yes, I need someone [14:21:52] 07/10/2012 - 14:21:51 - Created a home directory for jan in project(s): bastion [14:22:17] Jan_Luca: What would you like the name of the project to be? [14:22:33] wikiversity-sandbox [14:22:50] 07/10/2012 - 14:22:50 - User jan may have been modified in LDAP or locally, updating key in project(s): bastion [14:23:52] 07/10/2012 - 14:23:52 - Creating a project directory for wikiversity-sandbox [14:23:52] 07/10/2012 - 14:23:52 - Created a home directory for jan in project(s): wikiversity-sandbox [14:24:11] 07/10/2012 - 14:24:11 - Updating keys for jan at /export/keys/jan [14:24:31] Jan_Luca: So, you'll want an instance running mediawiki, right? Do you mind being my guinea pig? I recently set up some automation for that but no one but me has tested it. [14:24:51] 07/10/2012 - 14:24:51 - User jan may have been modified in LDAP or locally, updating key in project(s): wikiversity-sandbox [14:25:31] I can be your guinea pig [14:27:05] cool :) Before you create an instance, you'll want to set up a security group to allow web access. So... visit this page: https://labsconsole.wikimedia.org/wiki/Special:NovaSecurityGroup [14:27:38] Jan_Luca: I think you should be able to see the web security group for other projects there, and you can duplicated it for your project. If that isn't obvious let me know. [14:28:24] I only see bastion and our project [14:29:19] I only see bastion and our project [14:29:21] At the top of the page there's a 'show project filter' link, that'll let you view other projects (I think). If mwreview is in that list then look at that. [14:29:24] #Wikimania Ryan_Lane [14:30:14] There are only these two projects in my filter [14:30:19] Jan_Luca: I'm not 100% sure how project membership affects views of that page. So if you don't see what I see that's not necessarily surprising. [14:30:31] Ah, ok :) [14:30:56] In that case... do you want me to talk you through making the security group, or would you prefer I just make it for you? [14:31:42] I prefer you would do that :-) [14:32:32] ok, give me a minute or two. [14:32:46] ok [14:33:11] Ryan_Lane, Ryan_Lane1: Did you or someone else follow up with Sumana regarding "please give Labs account creation powers to a few other folks"? [14:33:24] andrewbogott: kind of [14:33:34] I'm at the hackathon and am giving it to them here [14:33:58] That works :) I was also going to offer to lurk on IRC and help folks, but you being on site is much better. [14:34:25] heh [14:34:49] I was busy when someone wanted irc help so I told them to mail the list... then replied via email :( [14:35:02] I feel horrible :( [14:35:14] Jan_Luca: I need to add myself to your project to tinker with it, I'll remove myself once things are squared away. [14:35:48] no problem [14:35:52] 07/10/2012 - 14:35:52 - Created a home directory for andrew in project(s): wikiversity-sandbox [14:35:56] Ryan_Lane: You and Sumana are Wikimania plague-bearers. After the conference, infected wikipedians will scatter to the four corners of the earth, civilizations will topple, etc. [14:36:47] Ryan_Lane: You're still sick from two weeks ago? Or just jetlagged? [14:36:56] 07/10/2012 - 14:36:56 - User andrew may have been modified in LDAP or locally, updating key in project(s): wikiversity-sandbox [14:39:34] hi andrewbogott: I'm also in need of a new project [14:40:14] GChriss: Ok, give me a few minutes to catch up [14:40:25] i can do that [14:40:28] no prob [14:40:36] mutante: better yet [14:40:40] GChriss: what's the project name you desire? [14:41:44] metavidwiki, (context: https://labsconsole.wikimedia.org/wiki/Requests#MetaVidWiki ) [14:41:46] Jan_Luca: Apologies, labsconsole page loads are very slow today due to wikimania so every little setting change takes minutes vs. seconds [14:42:24] no problem [14:42:36] I'm already in the bastion group [14:42:36] I have time [14:42:52] 07/10/2012 - 14:42:52 - Created a home directory for jasonspriggs in project(s): bastion [14:42:54] GChriss: user name gchriss? [14:43:25] yes [14:43:53] 07/10/2012 - 14:43:52 - Creating a project directory for metavidwiki [14:43:53] 07/10/2012 - 14:43:53 - Created a home directory for gchriss in project(s): metavidwiki [14:44:31] !log bastion created new project metavidwiki with gchriss as admin [14:44:34] Logged the message, Master [14:44:35] GChriss: done [14:44:46] much appreciated :) [14:44:52] 07/10/2012 - 14:44:52 - User gchriss may have been modified in LDAP or locally, updating key in project(s): metavidwiki [14:45:23] !log metavidwiki added gchriss as netadmin and sysadmin [14:45:25] Logged the message, Master [14:45:28] ^logging example [14:47:45] Jan_Luca: OK, now you have a security group called 'web'. Instances will need to be in that group in order for you to access web pages on those instances. [14:48:14] So, you can go ahead and create an instance now. Best to use Precise, and some size bigger than 'tiny'. [14:48:16] !log bastion labs now has > 100 projects - We just created project #101. [14:48:18] Logged the message, Master [14:48:49] 2 Questions: Does everybody need a labs account to ssh an instance [14:49:10] And when I need a dns domain I sould ask here? [14:49:36] GChriss: what andrew just said to Jan_Luca also applies to you:) [14:49:48] Jan_Luca: Yes, people will need labs accounts to access the instance. [14:50:01] mutante: Except GChriss probably doesn't have a 'web' security group yet. [14:50:10] GChriss: well, i did not create a security group. But the part about creating instances [14:50:15] Jan_Luca: Public IPs are in short supply, so it might be a struggle to get you one. [14:50:16] andrewbogott: yes [14:50:57] and dns names [14:50:58] ? [14:51:18] Jan_Luca: In theory we will eventually have a proxy system so that you can provide public web access to an instance without a public IP. But we haven't made much progress there. [14:51:52] Jan_Luca: Yeah, at the moment the dns question is the same as the public ip question (that is, a dns record doesn't help unless you have a public IP to point it at.) [14:52:23] Jan_Luca: you can already proxy through the bastion host though, you can use the ssh connections as a SOCKS proxy in your browser [14:52:36] Would be awesome to get a service written up for proxying stuff rather than just installing a couple of nginx boxes with pacemaker or such (though due to ebtables we'd probably have to ask the nova api nicly to re-assign the ip over just arping for it...) [14:52:54] mutante: But that only works for users who already have a bastion login, right? [14:53:17] andrewbogott: true. i assumed that if he has his own project [14:53:32] yep. Just clarifying 'access for Jan' vs. 'access for the world' [14:53:46] 'access for the world' might be a bad idea anyway, with current labs hardware. [14:53:47] true [14:54:14] You just need a large trout to keep wikipedia users in check with complaining :D [14:54:28] Jan_Luca: Following all this? [14:54:33] I am [14:54:34] yes [14:55:12] Jan_Luca: Go ahead and set up a 'naked' instance right now. We'll make sure that it's up and working and you have ssh access, and then we can get mediawiki up. [14:55:22] But can I use more than 1 dns with one IP? [14:55:30] It will probably take ~20 minutes for the instance to build. [14:55:39] Is it the hackathon over the next few days? [14:55:40] Jan_Luca: Yeah, that's possible. [14:55:59] would it make sense to start something of a waiting list for IP addresses? to track/prioritize requests? [14:56:19] I, um, need an IP address [14:57:09] there's [[Requests]] [14:57:28] GChriss: Probably worth asking Ryan_Lane about that. I think that there are so few IP addresses available that basically no one gets them. [14:57:39] A proxy solution is really what we need. [14:58:19] But, yeah, in the meantime, some kind of policy for deciding who gets them would be useful (even if that policy is "no one gets them") [15:01:19] GChriss: hmm, yea maybe just dump it in Requests/IP for now and then mention it on list, [15:01:27] * Damianz wonders how many broken vms we still have [15:01:49] GChriss: btw, you saw this yet? https://labsconsole.wikimedia.org/wiki/Nova_Resource:Metavidwiki [15:02:24] GChriss: that's like your project home page where the logs end up and you see your members, monitoring links etc.. [15:03:10] * Damianz thinks labs nagios isn't here... that's why it's really not spammy [15:03:28] GChriss: so doing !log metavidwiki requesting IP ... would also kind of make sense i guess [15:03:58] Jan_Luca: Instance creation going OK? [15:04:15] Damianz: the bot might be down, the Nagios web ui is up though it looks [15:04:23] https://labsconsole.wikimedia.org/wiki/Nova_Resource:Wikiversity-sandbox show it [15:05:31] mutante: Yeah, I don't think it's running under upstart/supervisor anymore since it got rebuilt. Was spamming a lot hence blocked but I was hoping it was going to behave if people are activly working on stuff and want to know it's broken :) [15:05:47] Damianz: hmm, i see, it's kind of up, but also broken http://nagios.wmflabs.org/cgi-bin/nagios3/status.cgi?hostgroup=deployment-prep&style=detail&hoststatustypes=2 [15:05:59] Jan_Luca: Meanwhile, let's see if you can ssh into bastion. [15:06:02] Deployment is a little screwed still :( [15:06:19] I think 30/31 need re-installing but hashar would know best [15:06:32] busy rebasing, sorry [15:06:34] * Damianz hopes they are mostly puppetized apart from squid [15:06:37] beta is screwed yes [15:06:48] would work on it as soon as possible [15:07:50] It sorta works but I'm assuming if I login and skip squid it will be broken... might see if I can get some time to fix at least a couple of apache boxes later, think I can remember how they are setup. [15:09:19] Sorry internet problems [15:10:04] I can log in into bastion.wmflabs.org [15:11:07] hashar: sorry about beta :( [15:11:17] Ryan_Lane: well shit happens :-D [15:11:38] Ryan_Lane: meanwhile, your reported apache30 / 31 to be screwed but they were still up and serving page on Sunday [15:11:41] mutante: yes, got it. I'll probably have questions later, but good for now [15:11:44] Ryan_Lane: are they just slightly corrupted? [15:11:47] hashar: yes [15:11:58] some instances still run [15:12:03] others are totally fucked [15:12:08] this is why I didn't notice it at first [15:12:08] GChriss: ok, great [15:12:13] Ryan_Lane: most of them I can reinstall them with puppet [15:12:15] but "ssh wikiversity-sandbox-web.pmtpa.wmflabs" does not work [15:12:25] Ryan_Lane: I guess they will land on one of the new servers? [15:12:31] yep [15:12:39] which is good [15:12:41] Jan_Luca_: you need a ProxyCommand in your sshconfig [15:12:55] the new servers don't have problems. just the migration stuff is broken [15:13:05] on bastion? [15:13:05] Ryan_Lane: the only thing I can blame you on is attempting a migration on a Friday ;-] [15:13:09] heh [15:13:10] true [15:13:12] Ryan_Lane: Is it possible to re-launch a server with the same settings in console? IE without having to copy everything over? Would be useful [15:13:21] Damianz: not really [15:13:29] It's a feature I plan on adding [15:13:42] "saved profiles" or something like that [15:13:46] 'Clone' or suck would be useful for so many things, for now by hand will do [15:14:03] Ryan_Lane: will definitely keep you up to date. I was planning to do fix beta in the evening, but no luck with my little daughter :-] [15:14:28] We could let Ryan baby sit at the hackerthon? :D [15:14:57] Jan_Luca: From bastion it should be trivial to ssh to your instance, but it maybe doesn't have sshd up yet. Lemme look. [15:15:06] !log metavidwiki requesting public IP address on [[Requests]] [15:15:07] Logged the message, Master [15:15:38] Ryan_Lane: re: requests for public IPs (currently). people are aware we currently don't have many, should we log the requests anyways? [[Requests/IP]] or something? list? [15:16:03] Jan_Luca: I think it just isn't up yet. [15:16:25] How can I see this? [15:17:09] You could read up on proxy connections while you're waiting. https://labsconsole.wikimedia.org/wiki/Help:Access [15:17:20] Jan_Luca: I just looked at the instance log, it seems busy. [15:17:34] um... 'get console output' on the instance page [15:18:43] thank you [15:20:06] Another question: Are server like apache or mysql are installed in each project speratly or are there central ones? [15:21:00] Jan_Luca: Per-instance, at the moment. [15:21:08] Ok [15:21:26] One your instance is up, I'd like you to try following this guide: https://labsconsole.wikimedia.org/wiki/Help:InstanceConfigMediawiki [15:23:40] *once [15:24:39] I will try it [15:28:11] 07/10/2012 - 15:28:11 - Updating keys for bkaempgen at /export/keys/bkaempgen [15:31:11] 07/10/2012 - 15:31:11 - Updating keys for bkaempgen at /export/keys/bkaempgen [15:32:29] Ryan_Lane: what does it mean when new instance logs look like this? http://pastebin.com/jnzxJN1E Is that somehow correct behavior? [15:37:50] Jan_Luca: I think you should force a reboot of your instance; seems to be misbehaving. [15:38:50] (And, if rebooting doesn't help, maybe start over. A surprisingly high % of new instances never make it, for reasons I've never understood.) [16:06:13] 07/10/2012 - 16:06:13 - Creating a home directory for karima at /export/keys/karima [16:07:12] 07/10/2012 - 16:07:12 - Updating keys for karima at /export/keys/karima [16:16:12] 07/10/2012 - 16:16:12 - Creating a home directory for kim at /export/keys/kim [16:17:12] 07/10/2012 - 16:17:11 - Updating keys for kim at /export/keys/kim [16:18:24] haha, i like this: "Preferred shell name" and somebody put in "/bin/bash" [16:18:32] rofl [16:18:33] it's correct, "shell user name" [16:18:46] should be in the template [16:37:23] ñMy project only had the default security group, so I created a web security group. How should I specify the security rule so I can access the web server using port forwarding? [16:44:12] slevinski: port: 80 tcp 0.0.0.0/0 would allow any to 80 [16:44:32] thanks [16:44:47] slevinski: there is a known issue with changing security group after instance creation. you might want to add the rules to your default group [17:03:22] Ryan_Lane: ping for another person needing gerrit approval? [17:03:35] sure. who's that? [17:03:51] fran [17:04:01] Fran McCrory [17:04:23] I mean, is there a link to the developer access request? [17:04:36] I need a set of info to create the account [17:04:54] It's on the developer access page, iirc. [17:04:57] Let me get the url [17:06:01] https://www.mediawiki.org/wiki/Developer_access#User:Fran_McCrory [17:06:01] Ryan_Lane: Do you know what the deal is with instances that do this? http://pastebin.com/jnzxJN1E [17:06:02] http://www.mediawiki.org/wiki/Developer_access#User:Fran_McCrory [17:06:04] heh [17:06:06] It happens rather frequently [17:06:17] that's normal [17:06:38] dhcp has a really short renew time [17:06:45] really? That instance has been buidling for ~90 minutes and I can't ping it. [17:07:10] which instance is it? [17:07:23] wikiversity-sandbox-web [17:07:43] But, ok, you're saying that the dhcp thing is unrelated to the failure to launch, which I can believe. [17:07:51] * Ryan_Lane nods [17:07:55] It's probably just that I only look at the logs of systems that are broken and always see that :) [17:07:56] lemme look at the console output and such [17:08:03] Amgine: done [17:09:26] andrewbogott: it looks deleted to me [17:09:31] https://labsconsole.wikimedia.org/wiki/Nova_Resource:I-00000312 [17:13:29] Ryan_Lane: Oh, ok. I told Jan that it was a lost cause and to delete it, he must've already done so. [17:13:40] So, the mystery remains... I'll bother you again the next time it happens. [17:13:41] ah [17:13:44] * Ryan_Lane nods [17:15:03] The next time it happens is there [17:15:12] Jan_Luca: where? [17:15:20] can you give me the instance name? [17:15:23] I cannot log in the new instance,t too [17:15:35] i-00000315 [17:15:48] wikiversity-sandbox-web.pmtpa.wmflabs [17:16:39] ok. gimme a sec to look at it [17:16:46] thank you [17:17:59] hm. network isn't working [17:18:01] strange [17:18:10] especially since it built [17:18:30] yep, same behavior as the last one. [17:18:41] this is on virt8 [17:19:21] Is the issue just that it needs to be in security groups default /and/ web, and it's just in web? [17:19:28] oh [17:19:28] yes [17:19:30] that's the problem [17:19:43] default should never be unchecked [17:19:44] ever [17:19:47] Jan_Luca, sorry, I misled you. [17:20:04] heh [17:20:08] Or, rather, didn't think to lead you as specifically as I should've. [17:20:09] documentation says that too :) [17:20:33] which, btw, I don't know that it isn't in the default group, just guessing. [17:20:47] yep, confirmed. [17:20:58] the default group has rules for ping and ssh [17:21:26] I had read the docu very well, my fault, too [17:21:32] heh [17:21:42] glad to know it's not an issue on virt8 :) [17:21:51] ok I create the instance new [17:32:41] !log deployment-prep creating apache32 and apache33 to replace the corrupted apache30 and apache31 instances [17:32:43] Logged the message, Master [17:47:34] now the ssh via i-00000317.pmtpa.wmflabs works but when I call on bastion "ssh wikiversity-sandbox-web.pmtpa.wmflabs" [17:47:49] I get "ssh: connect to host wikiversity-sandbox-web.pmtpa.wmflabs port 22: No route to host" [17:47:54] Jan_Luca: because it's named the same thing [17:47:59] it's cached in DNS [17:48:18] ill purge it for you [17:49:17] Jan_Luca: it's probably there [17:49:18] err [17:49:23] meaning, it should work now:) [17:52:12] thank you now it works [17:52:56] yw [18:00:34] Ryan_Lane: subscribe to what now? [18:00:49] 07/10/2012 - 18:00:49 - Created a home directory for jesusaurus in project(s): bastion,puppet-cleanup [18:00:52] jesusaurus: https://lists.wikimedia.org/mailman/listinfo/labs-l [18:00:59] and yes. it would be very ideal to turn each manifest into its own module [18:01:49] * Ryan_Lane nods [18:01:50] 07/10/2012 - 18:01:50 - User jesusaurus may have been modified in LDAP or locally, updating key in project(s): bastion,puppet-cleanup [18:02:34] uh oh, i cant clone the repo using ssh [18:02:36] that's more of a goal right now then ensuring puppet classes can cleanly run on first attempt [18:02:41] jesusaurus: it's git ;) [18:04:04] jesusaurus: https://labsconsole.wikimedia.org/wiki/Git [18:04:15] it shows how to clone and such [18:04:54] yes, im familiar with git, and have added my pubkey, but im getting permission denied [18:05:06] oh [18:05:11] you need to add it to gerrit too [18:05:18] it's separate for now [18:05:22] we have an open bug for that [18:07:46] ohh [18:08:24] yeah. it's annoying [18:08:52] 07/10/2012 - 18:08:52 - User jesusaurus may have been modified in LDAP or locally, updating key in project(s): bastion,puppet-cleanup [18:09:12] 07/10/2012 - 18:09:12 - Updating keys for jesusaurus at /export/keys/jesusaurus [18:11:21] so, what version of puppet are we using? [18:11:46] you said we are moving to parameterized classes first, so it must be a fairly recent version [18:12:31] 2.7.6 [18:13:26] yeah, fairly new [18:15:09] and how is puppet configured? if we simply make a new directory 'modules' beside 'manifests' at the top level, will the modules be found? [18:15:31] hmm [18:15:55] or is it your plan to put the modules in the 'manifests' directory? [18:16:13] funny enough I've never used modules :D [18:16:14] im not quite sure how that second way would work out [18:16:28] * andrewbogott thinks new modules should just go in the top level, maybe. [18:16:51] hm... maybe not. [18:16:58] well, we don't have the module path set [18:17:00] in the config [18:17:09] okay, so, what im used to is having a modules directory beside manifests, then inside the modules directory, each module has its own directory named after it [18:17:39] so the init.pp for the foobar module would be at ./modules/foobar/manifests/init.pp [18:17:42] Defaults to /etc/puppet/modules:/usr/share/puppet/modules [18:17:58] * Ryan_Lane nods [18:18:00] and foobars templates would be in ./modules/foobar/templates/something.erb [18:18:06] yep [18:18:29] so, we should add a modules directory to the repo [18:18:35] and slowly move them across [18:18:45] mhmm [18:18:47] https://labsconsole.wikimedia.org/wiki/Help:SelfHostedPuppet [18:18:57] you'll want to use puppetmaster::self to test your changes [18:19:28] once everything is a module dealing with puppet in labs will be easier [18:20:16] okay, i need to get back to work, but i'll start working on this some tonight [18:20:35] awesome. thanks for the help [18:20:39] let me know if you need anything [18:21:09] will do [18:21:10] jesusaurus: I'm interested in this undertaking too, feel free to delegate tasks to me. [18:21:20] andrewbogott: great [18:21:33] paravoid is likely wanting to do this as well [18:21:46] i need to look over the current codebase to figure out which modules will be easy to start with ;) [18:22:09] gerrit would likely be a good one: https://gerrit.wikimedia.org/r/#/c/13484/5 [18:22:26] ^demon started modifying that one to suck less [18:22:28] and r0csteady wants to help too, right r0csteady? [18:22:28] hey does anyone need me? [18:22:49] petan2: dunno, I think hashar started on fixing beta [18:23:11] ok [18:23:11] jesusaurus: you can add r0csteady to the puppet-cleanup project [18:23:14] since you're in it [18:23:16] I created a few instances, waiting for them to get installed [18:23:18] oh, cool [18:23:22] then will add the puppet classes [18:23:33] ah, I forgot to add you as sysadmin/netadmin [18:23:45] petan2: you could cleanup the deployment-bastion instance. I think it never worked and is currently corrupted according to Ryan mail [18:23:57] deployment-feed I can't remember what there was on it :/ [18:24:01] jesusaurus: you can add him to bastion, as well [18:24:07] ok [18:24:54] I'll definitely work on modules next week [18:24:56] when I get back [18:25:08] wow. wtf: http://ganglia.wikimedia.org/latest/?c=Virtualization%20cluster%20pmtpa&h=virt7.pmtpa.wmnet&m=load_one&r=hour&s=by%20name&hc=4&mc=2 [18:26:00] similarly: http://ganglia.wikimedia.org/latest/graph_all_periods.php?h=virt8.pmtpa.wmnet&m=load_one&r=hour&s=by%20name&hc=4&mc=2&st=1341944720&g=cpu_report&z=large&c=Virtualization%20cluster%20pmtpa [18:26:11] it must be waiting on the damn nfs server [18:26:22] I really need to spend some time to kill that off [18:29:20] andrewbogott: I have installed role::mediawiki-install::labs [18:31:11] Jan_Luca: Did you force a puppet run? Are things working at all? [18:31:42] I force a run. [18:31:46] How to test it? [18:32:44] as root: puppetd -tv [18:33:42] Jan_Luca: You'll need to set up a web-proxy somehow (e.g. https://labsconsole.wikimedia.org/wiki/Help:Access#Accessing_web_services_using_a_SOCKS_proxy) [18:34:05] Once you have web access to that machine, then there should be a mediawiki server at /wiki. [18:37:13] petan2: Sven_Manguard would like some help with bots [18:37:20] petan2: are you at wikimania? [18:40:13] Moment I start a virtual machine for the proxy [18:44:02] andrewbogott I'm away for 30 minutes [19:02:08] !log deployment-prep running puppets -tv on apache32 and apache33. Should make them able to serve Apache traffic after reboot. [19:02:10] Logged the message, Master [19:02:14] Jan_Luca: That puppet class seems to have regressed :( I'm working on it, but if you have other useful things to do you shouldn't wait on me. [19:13:35] andrewbogott: So I'm back now [19:13:51] Do the class install already apache, php and mysql? [19:14:07] It should, not sure if it did. [19:15:26] Apache it installed [19:16:09] PHP and mysql too [19:17:34] the code of MediaWIki is in /srv and not in /src/mediawiki like described [19:17:42] maybe this is the problem? [19:22:04] It should be linked I think. I'm setting up a dev box to investigate. [19:22:42] !log deployment-prep rebooting apache32 and apache33 (puppet run finished) [19:22:43] Logged the message, Master [19:23:08] there is a link /var/www/srv/mediawiki -> /srv/mediawiki [19:23:19] but there is no /srv/mediawiki [19:23:56] the files of mediawiki are in /srv [19:38:15] !log wikiversity-sandbox Created instance "wikiversity-sandbox-web" as web server instance [19:38:16] Logged the message, Master [19:39:45] Jan_Luca: I suspect that the git:clone base class changed its behavior. Should have a fix pretty quickly. [19:55:28] andrewbogott: When I changed the settings in /etc/apache/sites*/wiki: [19:56:22] to "Alias /w /srv/mediawiki" + "Alias /wiki /srv/mediawiki/index.php" [19:56:26] I get the wiki [19:56:53] so you only have to get the files into /srv/mediawiki to get this working [19:57:35] In theory you should get a preconfigured wiki. It's probably too late to make that happen on your instance though. [19:58:11] ok that has not work I get the "Please set up the wiki first."-page [20:00:17] Anyone here from BRFA on En.Wiki? Please ping me. [20:04:53] andrewbogott: Do you have decide where to put a request for a public IP on? [[Request]], [[Request/IP]] or something else? [20:06:02] Ryan_Lane: Suggestions? ^^ [20:19:58] Jan_Luca: OK, that puppet class should be fixed. I had to run it a second time to get everything in place, but... pretty close. [20:20:10] Jan_Luca: You're welcome to start a fresh instance or just live with what you've got :) [20:20:32] It's somewhat possible that subsequent puppet runs on that instance will rearrange things a bit, so maybe force a puppet run now to make sure it doesn't get too broken. [20:22:31] I recreate the instance [20:23:53] 07/10/2012 - 20:23:53 - Creating a project directory for linkcache [20:23:54] 07/10/2012 - 20:23:53 - Created a home directory for laner in project(s): linkcache [20:24:53] 07/10/2012 - 20:24:53 - User laner may have been modified in LDAP or locally, updating key in project(s): linkcache [20:25:06] Jan_Luca: either send a mail to the list or put in a bug [20:25:36] ok [20:32:18] Jan_Luca: I hope things go smoothly this time :/ [21:07:32] andrewbogott: I've started a puppet run. I'm now for a 30+ minutes away. When I'm back I will tell you the result [21:08:06] ok! Probably you'll need to do a second run to get things into a stable state. Something is out of order in my manifest. [21:10:27] ok [21:50:56] andrewbogott: I start the run a second time [22:01:03] andrewbogott: You can find the installed pupet without any change here: http://sandbox.wikiversity.wmflabs.org/wiki [22:01:14] It seems to work [22:01:20] Hey, it's up! [22:01:38] That's great. Thanks for your patience -- I hope that turns out to have been useful to you. [22:01:53] Yes, it was! [22:02:03] But I go to bed now. [22:02:19] A reasonable choice at any hour. [22:02:26] catch you later! [22:02:32] Bye! [22:46:16]