20:02:04 <mmcgrath> #startmeeting
20:02:04 <zodbot> Meeting started Thu Oct 22 20:02:04 2009 UTC.  The chair is mmcgrath. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:02:04 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
20:02:11 <mmcgrath> #topic Who's here?
20:02:12 <smooge> here
20:02:24 * a-k is
20:02:54 * mdomsch ducks
20:03:20 <mmcgrath> Well, lets just get started.
20:03:25 <mmcgrath> #topic -- Beta Release
20:03:51 <mmcgrath> https://fedorahosted.org/fedora-infrastructure/report/9
20:03:53 <skvidal> hi
20:04:05 <mmcgrath> Looks like almost everything was closed
20:04:06 <mmcgrath> .ticket 1711
20:04:07 <zodbot> mmcgrath: #1711 (Release day ticket) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/1711
20:04:09 * mmcgrath closes that now.
20:04:26 <mmcgrath> mchua: you going to be around for a bit?
20:04:29 * jcollie is here for a few minutes till my plane boards
20:04:37 <mmcgrath> jcollie: k.
20:04:39 <mmcgrath> .ticket 1715
20:04:40 <zodbot> mmcgrath: #1715 (Lessons Learned) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/1715
20:04:47 <mmcgrath> So the beta launch went off without a hitch as far as I know.
20:04:54 <mmcgrath> Anyone think otherwise?
20:05:05 <mmcgrath> I guess there was one ppc link missing from the get-prerelease page.
20:05:16 <mmcgrath> The mirrors were in great shape by the time we sent the email out
20:05:23 <mdomsch> how early was the bitflip?
20:05:30 <mmcgrath> 4 hours prior to the release
20:05:33 <mdomsch> fwiw, the crawler is taking ~4 hours to get through everyone
20:05:47 <mmcgrath> At the time of release we were at 85% success rate.
20:06:01 <mmcgrath> I actually never saw any failures while browsing but our test script found some.
20:06:17 * nb here
20:06:47 <mmcgrath> Anyone have anything else on the release?
20:07:20 <mmcgrath> k
20:07:24 <mmcgrath> in the interest of jcollie's plane
20:07:33 <mmcgrath> #topic FAD for fedoratalk starts tomorrow
20:07:43 <jcollie> woo baby
20:07:46 <mmcgrath> jcollie: you have any last minute details you need me to take care of or are we in good shape?
20:08:11 <jcollie> have the asterisk 1.6.2 rpms been copied to the infra repos?
20:08:21 <jcollie> they are in my puppet1 home dir
20:08:31 <mmcgrath> jcollie: doesn't look like it, or if they have they didn't get in the right repos
20:08:36 <mmcgrath> 1.6.1.6-2 is still on asterisk2.
20:08:38 * ricky is here
20:08:40 <mmcgrath> I'll get that right after the meeting.
20:08:54 <jcollie> that's about it as long as asterisk2 is in good shape i think we're good for tomorrow
20:08:58 <smooge> there were some errors in puppet after I removed the lock on asterisk2
20:08:59 <mmcgrath> Cool.
20:09:05 <mmcgrath> smooge: yeah I just disabled it again
20:09:12 <mmcgrath> I'll get it in shape so there are no errors by tomorrow.
20:09:55 <smooge> i noticed that there was an article on LWN today about a hardware asterisk unit that was 'completely' open
20:10:05 * mmcgrath saw that too.
20:10:11 <mmcgrath> +1 to open hardware :)
20:10:16 <mmcgrath> jcollie: ok, anything else besides that?
20:10:19 <jcollie> that'd be cool... is the LWN article in the open yet?
20:10:25 <jcollie> i don't think so
20:10:30 <mmcgrath> jcollie: http://lwn.net/SubscriberLink/357809/eb0646c5d4839517/
20:10:41 <smooge> you beat me
20:10:55 <mmcgrath> Ok, so that's really it on the FAD which starts tomorrow!
20:10:58 <jcollie> hmm i'll have to wait a couple weeks to read it i don't have a sub
20:11:01 <mmcgrath> #topic Fedora Insight
20:11:04 <mmcgrath> mchua: ping
20:11:54 <mmcgrath> ok, I'll assume mchua's out for now
20:12:04 <mmcgrath> I've got the Fedora Insight staging install and going (at least what I could)
20:12:06 <nb> I think all of the packages for FI have been reviewd and approved
20:12:14 * nb just approved xinha a few minutes ago
20:12:16 <mmcgrath> the only thing missing is the branding stuff.
20:12:31 <mmcgrath> Which is why it looks like this atm - https://stg.fedoraproject.org/insight/
20:12:44 <mmcgrath> shoot, still no J5 :)
20:12:47 <mmcgrath> that's his ear :)
20:13:05 <mmcgrath> Eventually it will look like this - http://publictest6.fedoraproject.org/zikula/
20:13:10 <mmcgrath> mizmo: you around by chance?
20:13:21 <mizmo> mmcgrath: yeppers
20:13:27 <smooge> cool
20:13:36 <smooge> to the look
20:13:40 <mmcgrath> mizmo: oh!  actually I may have answered my own question.
20:13:43 <mizmo> kk
20:13:51 <nb> nice, i like what it will look like
20:13:55 <mmcgrath> mizmo: I noticed none of the pictures were sized correctly, then I realized they were stored elsewhere
20:14:01 <mmcgrath> like MEGA J5 and MONDO lmacken
20:14:03 <mmcgrath> http://duffy.fedorapeople.org/temp/fc-photos/IMG_0743.JPG
20:14:21 <nb> #link https://stg.fedoraproject.org/insight/ - Fedora Insight Staging Instance
20:14:21 <mizmo> mmcgrath: yep zikula doesn't have any workflow for images that i could find
20:14:22 <smooge> though the speed of that site is slow...
20:14:26 <mmcgrath> J5's head is so big I can see what he's thinking.
20:14:27 <abadger1999> mmcgrath: question --  a value add or will it be frozen for release?
20:14:43 <nb> #link http://publictest6.fedoraproject.org/zikula/ - PT Fedora Insight Instance
20:14:44 <mmcgrath> abadger1999: I believe won't get frozen.
20:14:47 <mmcgrath> not in the critical path.
20:14:49 * jcollie is outta here until I get to chi-town
20:14:55 <abadger1999> Cool.
20:14:56 <smooge> see you later jc
20:15:23 <mmcgrath> abadger1999: I'll make a note to ensure our docs are all up to date.
20:16:19 <mmcgrath> Ok, so there's really nothing else on this atm.  I need to talk to mchua and make sure everything works out as it should
20:16:38 <mmcgrath> next
20:16:40 <mmcgrath> a-k: ping
20:16:42 <mmcgrath> #topic Search Engine
20:16:47 <mmcgrath> a-k: want to take this one?
20:16:51 <a-k> Sure
20:17:02 <a-k> I'll remind everyone that the wiki page for search is http://fedoraproject.org/wiki/Infrastructure/Search
20:17:14 <a-k> We've moved the C/C++ engines from "not suitable" to "in progress" but kept the Java ones in the "not suitable" pile.
20:17:24 <a-k> That's 10 candidates (perl and C) overall.
20:17:35 <a-k> I've started looking at KinoSearch, which is a perl/C hybrid that keeps indexes in db3 files.  huzaifas is still looking a Namazu, which I think is perl only.
20:17:50 <a-k> At this pace, it'll take us a little over a month to get through them all.
20:17:56 <mchua> mmcgrath: pong. *reads backlog*
20:18:10 <mmcgrath> mchua: stick around for a bit if you have time, we'll chat after a-k's done
20:18:13 * mchua is actually in the middle of FUDCon meeting atm, but will catch up on backlog real quick.
20:18:14 <lmacken> I've got another engine candidate that a couple of my college buddies wrote, which I will be using for a project of mine: http://code.google.com/p/logstash/ (db4 storage)
20:18:35 <mmcgrath> a-k: ^^ add it to the list
20:18:43 <a-k> lmacken:  Okay.  I'll have a look.
20:19:05 <mmcgrath> a-k: what have your thoughts been so far?  Pretty confident something is out there for us or will we have to lower our requirements a bit?
20:19:07 <lmacken> I'll be deploying it probably this weekend or next week, so I'll let you know how it goes
20:19:17 <a-k> I think that's all I planned to say.
20:19:20 <lmacken> ruby app.. but JSON interfaces, so I'm happy
20:19:45 <a-k> I think there should be something we can use.
20:19:46 <mmcgrath> a-k: k, are you guys going to have your tests publically available or just using workstations and such?
20:19:59 * mmcgrath is kind of excited about this one
20:20:11 <a-k> I'm just using my workstation.  huzaifas is the same.
20:20:32 <a-k> When we decide we like something, that's when we would put it on a t machine.
20:20:39 <mmcgrath> a-k: k, when it gets to that point that you want to show this stuff off (and get size requiremnets) let me know and I'll make sure to get you two some publictest servers to use.
20:20:47 <mmcgrath> I suspect we might need a special one though.
20:20:59 <mmcgrath> I honestly have no idea how much actual searchable content we have, nor how much storage it will require
20:21:21 <mmcgrath> a-k: ok, anything else?
20:21:24 <a-k> Okay.  Mostly we're just trying the candidates out for size now, anyway.
20:21:31 <a-k> Nothing else, I think.
20:21:34 <mmcgrath> k
20:21:58 <mmcgrath> mchua: if you don't want to do double duity with your other meeting totally cool, we'll meet up when you have some more time.
20:22:20 <mchua> mmcgrath: I'll be out of that meeting in... probably 5, 10 minutes.
20:22:24 <mchua> It's a short one.
20:22:35 <mmcgrath> k
20:22:44 <mmcgrath> #topic PHX1->PHX2 move
20:22:56 <mmcgrath> So this keeps getting pushed back though now we're looking more solid for mid December.
20:23:00 <mmcgrath> the weekend after FUDCon actually.
20:23:18 <mmcgrath> The hardware we have in PHX2 now won't get installed until mid November.
20:23:24 <mmcgrath> so for now, there's really not much for us to do but plan.
20:23:38 <mmcgrath> as soon as that hardware gets installed though, I want to get our temporary stuff setup
20:23:49 <mmcgrath> particularly the databases.
20:24:19 <mmcgrath> smooge: you have anything else on this?
20:24:33 <mmcgrath> There were some talks about switching up our primary mirror system.
20:24:35 <mmcgrath> which we still might do
20:24:37 <mmcgrath> Oxf13: ping
20:24:39 <smooge> no I haven't heard anything. I need to get the latest docs to ping releng with
20:24:43 <mmcgrath> I forgot to mention this.
20:25:07 <mmcgrath> The master mirror in PHX2 might be brought online soon.
20:25:14 <smooge> ah
20:25:16 <mmcgrath> It was supposed to be up back in April or so but plans got switched around.
20:25:28 <mmcgrath> Now they were going to have us use it as the canonical mirror but I put a no on that until after F-12 ships.
20:25:30 <smooge> what will that look like to us
20:25:41 <mmcgrath> until the move, it'll look just like RDU to us
20:25:49 <mmcgrath> except on regular internet, not i2.
20:25:58 <mmcgrath> which is good in case our mirror gets overwhelmed during this time.
20:26:09 <mmcgrath> after the move though, it'll be the actual device we mount on /pub/
20:26:17 <mmcgrath> I gave them a cutoff date of November 1st.
20:26:25 <mmcgrath> If they can't get it done by then, they'll have to wait until after F-12 ships
20:27:24 <mmcgrath> So that's really all I have on that.
20:27:30 <mmcgrath> anyone have any questions or concerns?
20:28:02 <mmcgrath> 'mmm kay
20:28:18 <mmcgrath> #topic Zabbix
20:28:24 <mmcgrath> smooge: how's the zabbis stuff been going?
20:28:26 <mmcgrath> err zabbix.
20:28:47 <smooge> well i created a new rule and am trying to get it to do my other bidding
20:29:03 <smooge> it likes to say its removing a rule and then show it up again on refresh
20:29:09 <smooge> or not.
20:29:15 <mmcgrath> <nod>
20:29:20 <mmcgrath> so one thing I wanted to mention with zabbix...
20:29:28 <smooge> so I am reading through to figure what I needed to do right
20:29:36 <smooge> versus what I was doing :)
20:29:41 <mmcgrath> The dump size for zabbix is almost the size of smolt.
20:29:45 <mmcgrath> meaning... it's a very large database.
20:29:49 <smooge> ggruener has set up an excellent tool too
20:29:51 <smooge> ah
20:29:54 <mmcgrath> I'm thinking we're going to want to do something about that
20:29:56 <ricky> It's all InnoDB though :-)
20:30:00 <mmcgrath> either not keep data for as long, or not keep as much of it.
20:30:04 <mmcgrath> ricky: ha, is it?
20:30:14 <smooge> or run it on a seperate db?
20:30:15 <mmcgrath> I do see lots of zabbix queries show up in /var/log/mysq/slow_queries.log though
20:30:24 * ricky wonders how slow restores are, might be good to test on db1.stg if it has the space
20:30:46 <mmcgrath> smooge: possibly
20:30:47 <smooge> the question I will need to look is if zabbix is one of those "what mysql is used for other things than me?" tools or not
20:31:04 <mmcgrath> I think some of the problem are that we're storing minute level data for some stuff that probably doesn't change but once an hour or once every 6 hours.
20:31:15 <smooge> ah
20:31:27 <mmcgrath> Just something to look at.
20:31:34 <mmcgrath> The good news is it doesn't seem to really be causing any problems.
20:31:44 <mmcgrath> and having those extra checks on the clients doesn't seem to be impacting them.
20:32:27 <mmcgrath> Anyone have anything else on that?
20:32:31 <mmcgrath> just figured it'd be a good one to check up on
20:32:43 <ricky> Tiny question about the new puppetlock check
20:32:57 <ricky> Is it only checking for *empty* lock files?
20:33:10 <mmcgrath> ricky: I think it's checking age.
20:33:18 <mmcgrath> if the puppet lock file exists there for 6 hours, we want to know about it
20:33:20 <ricky> Ah, OK that works
20:33:21 <smooge> no its supposed to check thtat the file is over X seconds old
20:33:29 <Lilium> http://i34.tinypic.com/25gton5.jpg any idea why when i choose wlan1 (broadcom .. . . ) after i change the tab or click ok it switches back
20:33:31 <Lilium> ..
20:33:37 <smooge> I am not sure it does that 100% yet.
20:33:37 <Lilium> yea im new in this
20:33:44 <ricky> Lilium: Hi, try asking in #fedora
20:33:49 <Lilium> k
20:34:31 <mmcgrath> ricky: so the thought is, even when a normal puppet run is going and /var/lib/puppet//state/puppetdlock exists.
20:34:38 <mmcgrath> unless it's been there for 6 hours, we won't get notified about it.
20:34:51 <mmcgrath> So we don't need to check if it's empty or not
20:34:52 <ricky> Yeah, I see now
20:34:59 <mmcgrath> because if puppet has been running for 6 hours... something's broke :)
20:35:05 <mmcgrath> smooge: ^^
20:35:28 <smooge> yes.
20:35:37 <mmcgrath> Ok, if nothing else on that
20:35:40 <mmcgrath> #topic Open Floor
20:35:45 <mmcgrath> anyone have anything they'd like to discuss?
20:35:56 <smooge> nothing from me at the moment.
20:36:24 <nb> oh
20:36:33 <nb> blogs i think Ac-town is looking at it
20:36:39 <nb> he took the ticket a few days ago
20:37:10 <mmcgrath> nb: yeah, what's the poop on that?
20:37:13 <mmcgrath> we live yet or not?  :)
20:37:21 <nb> we are, but are not
20:37:23 <mmcgrath> we need to make sure all that stuff is getting backed up and stuff.
20:37:30 <nb> the ssl is causing issues again
20:37:44 <nb> i will email Ac-town and try to take a look at it tomorrow
20:37:53 <mmcgrath> nb: thanks
20:37:54 * ricky mentions that there's a new wordpress security update out
20:37:57 <mmcgrath> mchua: all done?
20:38:00 <skvidal> if anyone has a free cycle or two this might be worth pursuing: http://gobby.0x539.de/trac/ticket/57#comment
20:38:02 <ricky> For a DOS
20:38:21 <mmcgrath> .whoowns gobby
20:38:22 <zodbot> mmcgrath: lmacken
20:38:33 <nb> dnssec is basically done. i'm just getting them all in dlv.isc.org now
20:38:38 <mmcgrath> skvidal: is that something that we'll just get with the next version?
20:38:49 <mmcgrath> I mean, it'll need configuration and stuff.
20:38:49 <skvidal> mmcgrath: it'll need to get tied into fas, I suspect
20:38:54 <mmcgrath> yeah
20:38:56 <skvidal> right
20:39:50 <smooge> nb, thanks on the DNS-sec stuff
20:40:10 <nb> mmcgrath, i got your email, btw
20:40:52 <mmcgrath> nb: <nod>
20:41:33 <mmcgrath> Ok, anyone have anything else they'd like to discuss?
20:41:40 <abadger1999> mmcgrath, skvidal: BTW, that will also go into the infinote server, not sobby
20:42:05 <skvidal> oh - then I misread that
20:42:09 <mmcgrath> infininote?
20:42:12 * mmcgrath doesn't even know what that is
20:42:17 <mchua> mmcgrath: done!
20:42:19 <skvidal> it's the new gobby server
20:42:22 * mchua reads backlog for FI stuff
20:42:22 <skvidal> http://gobby.0x539.de/trac/wiki/Infinote
20:42:34 <skvidal> mmcgrath: sounds like what we need to transition to from sobby
20:42:35 <abadger1999> skvidal: I mean, the feature needs to beadded to infininote.
20:42:39 <mmcgrath> mchua: no need I'll catch you up :)
20:42:45 <mmcgrath> mchua: ok, stg.fedorahosted.org/insight/ <--
20:42:46 <abadger1999> skvidal: Not that it's done yet :-(
20:42:50 <mmcgrath> it's basically up(ish)
20:43:00 <mmcgrath> can you give it a look over, see whats busted (besides the theme) ?
20:43:00 <mchua> all caught up.
20:43:14 <mmcgrath> when are you going to need me to take my next step?
20:43:16 <skvidal> abadger1999: you're no fun, you know that? :)
20:43:21 <mchua> Yep, I saw the staging link earlier, it looks great but I haven't tested much yet... *reloadeth*
20:43:30 <smooge> mmcgrath, the insight servers images were really slow
20:43:47 <mchua> holy cow it's a REALLY BIG EAR.
20:43:59 <mchua> ...but at least it stays within its column now.
20:44:04 <nb> mchua, i just finished the xinha review
20:44:24 <mmcgrath> smooge: they're from fedorapeople.org and are gigantic pictures :)
20:44:24 <ricky> smooge: All of the images, or just the giant J5?  :-)
20:44:46 <mchua> mmcgrath: FAS auth appears to be not working
20:44:49 <mchua> nb: thanks!
20:44:57 <nb> no problem
20:45:17 * mchua mentally reviews the remaining "must do for FI" things
20:45:23 <mmcgrath> mchua: oh!  yes.
20:45:28 <mmcgrath> I never did get that working.
20:45:38 <mmcgrath> it worked before doing an import of the pt6 db, does not work now.
20:45:40 <smooge> can't find a big ear
20:46:11 <mchua> smooge: https://stg.fedoraproject.org/insight/ and scroll halfway down
20:46:46 <skvidal> nice
20:46:50 <mmcgrath> mchua: ok, I'll try to get auth working again, are you sure you don't just want to start over?
20:46:53 <skvidal> such nice hair
20:47:08 * mmcgrath isn't sure what you have in mind for getting this to prodcution
20:47:14 <mchua> mmcgrath: starting over is fine by me, whatever is easiest for you.
20:47:21 <smooge> ah ok I was looking at http://publictest6.fedoraproject.org/zikula/ which looks completely different
20:47:32 <mmcgrath> oh, well if it's just the same to you, starting over is much easier :)
20:47:40 * mchua is going to braindump the "what needs to happen before production" list, and then people can tell me how wrong I am
20:48:02 <mmcgrath> do you know when the theme will be done?
20:48:45 <mchua> mmcgrath: Nope. That's something we need to holler out on; since mizmo had to drop back, we don't have anyone dedicated to working on the theme, just various folks picking up on bits and snatches.
20:48:53 <mmcgrath> k
20:49:10 <mchua> mmcgrath: Simon said he would have time to sprint on it this weekend, though, so if that's the major thing to wrap up for production, we can ask him to pretty much take care of that
20:49:16 <mchua> and then I think just packaging will be left, which is easy.
20:49:25 <mchua> * finish theme
20:49:28 <mchua> * package theme
20:49:35 <mmcgrath> yep, those are biggies ;)
20:49:42 <mchua> * make https://fedoraproject.org/wiki/How_to_set_up_a_Zikula_sandbox#Fedora_Insight_specific_modules a complete and final list with statuses
20:49:57 <mchua> * push everything on that list through review and into repos (almost done, but it's been at that stage for weeksl)
20:50:19 <mmcgrath> mchua: sounds good to me
20:50:26 * mmcgrath has nothing else on this for now
20:50:26 <mchua> * migrate db
20:50:26 <mmcgrath> anyone else have anything?
20:50:30 <mchua> * get FAS to work
20:50:36 <mchua> mmcgrath: I think that's it; the rest is content.
20:50:41 <mmcgrath> <nod>
20:50:43 <mmcgrath> Allrighty
20:50:49 <mmcgrath> if no one has anything else we'll close the meeting in 30
20:50:56 <nb> mchua, any more reviews to do?
20:51:33 <mchua> nb: One, I think; I'll ping you in #fedora-devel when this is over.
20:51:56 <mmcgrath> allrighty :)
20:52:06 <mmcgrath> mchua: thanks for stopping
20:52:08 <mmcgrath> by :)
20:52:17 <mmcgrath> that's it!  Closing meeting
20:52:19 <mmcgrath> #endmeeting