infrastructure October 2008

infrastructure@lists.fedoraproject.org

37 participants
59 discussions

Outage Notification - 2008-10-05 07:00 UTC
by Mike McGrath 05 Oct '08

05 Oct '08

There will be an outage starting at 2008-10-05 07:00 UTC, which will last approximately 2 hours. To convert UTC to your local time, take a look at http://fedoraproject.org/wiki/Infrastructure/UTCHowto or run: date -d '2008-10-05 07:00 UTC' Affected Services: Buildsystem CVS / Source Control Database DNS Fedora Hosted Fedora People Fedora Talk Mail Mirror System Torrent Websites Unaffected Services: All Ticket Link: https://fedorahosted.org/fedora-infrastructure/ticket/875 Reason for Outage: Doing our monthly updates to our systems, there's a kernel update this time around and we'll be rebooting everything. In theory most systems won't go down at all. But many will be down for a few minutes at a time during our 2 hour outage window. Contact Information: Please join #fedora-admin in irc.freenode.net or respond to this email to track the status of this outage.

1 2

refresh_mirrorlist_cache traceback
by Ricky Zhou 05 Oct '08

05 Oct '08

Hey, we've been seeing refresh_mirrorlist_cache errors like this recently: Traceback (most recent call last): File "/usr/share/mirrormanager/server/refresh_mirrorlist_cache", line 41, in ? sys.exit(main()) File "/usr/share/mirrormanager/server/refresh_mirrorlist_cache", line 36, in main doit() File "/usr/share/mirrormanager/server/refresh_mirrorlist_cache", line 16, in doit populate_all_caches() File "/usr/share/mirrormanager/server/mirrormanager/mirrorlist.py", line 248, in populate_all_caches populate_directory_cache() File "/usr/share/mirrormanager/server/mirrormanager/mirrorlist.py", line 113, in populate_directory_cache if hcurl is None or len(hcurl) == 0: continue TypeError: len() of unsized object It looks like it's coming from line 113 of /usr/share/mirrormanager/server/mirrormanager/mirrorlist.py: if hcurl is None or len(hcurl) == 0: continue Should this be changed to just: "if not hcurl: continue" ? I think hcurl comes from a SQLObject query, so I'm not 100% sure. Matt, could you take a look at this? Thanks, Ricky

2 1

download.fp.o https?
by Matt Domsch 04 Oct '08

04 Oct '08

not sure if we care, but download.fp.o is using the plain fp.o https cert, not the wildcard cert. We only use download.fp.o to do redirects to (most likely non-https) mirrors, but it might still be good to be sure there's no MITM between user and our redirector. If so, we should be using the wildcard cert, yes? Thanks, Matt -- Matt Domsch Linux Technology Strategist, Dell Office of the CTO linux.dell.com & www.dell.com/linux

2 1

MirrorManager upgrades
by Matt Domsch 04 Oct '08

04 Oct '08

----- Forwarded message from Matt Domsch <Matt_Domsch(a)dell.com> ----- Date: Sat, 4 Oct 2008 16:37:12 -0500 From: Matt Domsch <Matt_Domsch(a)dell.com> To: mirror-list-d(a)redhat.com Subject: MirrorManager upgrades References: <20081003150210.GJ1105163(a)hiwaay.net> <20081003151427.GK1105163(a)hiwaay.net> <E9909A75A543064DB66E55B8E3BE41ECA22457(a)ausx3mps307.aus.amer.dell.com> <20081003173951.GA1222229(a)hiwaay.net> <20081003193533.GA16972(a)auslistsprd01.us.dell.com> <20081003201159.GC1222229(a)hiwaay.net> Status: RO On Fri, Oct 03, 2008 at 03:11:59PM -0500, Chris Adams wrote: > Once upon a time, Matt Domsch <Matt_Domsch(a)dell.com> said: > > Found "yet another bug". :-) > > > > Please run report_mirror one more time, and check again after the top > > of the next hour. > > Well, they all came back good about half an hour ago, but now I'm back > to only fedora-9/{i386,x86_64,ppc} come back with the preferred netblock > and my mirror in the list (the other repos give a valid list, just not > including my local mirror). Chris and I traded private mails after this, and I believe everything is working again as expected. It helps if I deploy the new version to all of the machines running in the cluster... If anyone else has trouble with MM, please let me know. FWIW, I've been working on an upgrade to mirrormanager, to allow it to expose the mirrorlist as a metalink file. Longer-term, yum is growing the capability to use a metalink file for the mirrorlist, which brings with it the ability to check checksums and signatures of the repository files (whats in the repodata/ directories). It's not looking like that feature will make Fedora 10, but at least the MM code server-side will be in place. One feature of metalinks is the ability for the user's client app to download multiple chunks of a large file from multiple servers in parallel. For the moment I've disabled this feature in the metalink files being published, by setting maxconnections=1, and by not publishing info about smaller "chunks" of the ISO files. I know "download accelerators" have been problematic for our mirrors, and I don't want to exacerbate the problems that adding metalinks might bring. If you want to see for yourself what is being published, take a normal mirrorlist URL, such as: http://mirrors.fedoraproject.org/mirrorlist?repo=rawhide&arch=i386 and use 'metalink' instead of 'mirrorlist'. http://mirrors.fedoraproject.org/metalink?repo=rawhide&arch=i386 Likewise, URLs to ISOs of the form http://mirrors.fedoraproject.org/mirrorlist?path=pub/fedora/linux/releases/… become http://mirrors.fedoraproject.org/metalink?path=pub/fedora/linux/releases/te… This shows a long list of mirrors, plus some metadata about them: * a preference which is just the mirrormanager randomization algorithm * country each mirror is in * multiple protocols, if a mirror provides >1. In the past, if both http and ftp were provided, only the http URL would have appeared in the mirrorlist. This allows client-side tools (not yum yet, but perhaps in the future - yum-ftponly plugin anyone?) to determine the best protocol they can use. * SHA1SUM for verification In addition, mirrors.fedoraproject.org now can serve its content via https, e.g. https://mirrors.fedoraproject.org/metalink?repo=rawhide&arch=i386 for added security (again, assuming your tools actually do certificate checking, which ATM yum does not). Thanks, Matt -- Matt Domsch Linux Technology Strategist, Dell Office of the CTO linux.dell.com & www.dell.com/linux ----- End forwarded message -----

1 0

smtp-server?
by Thomas Spura 03 Oct '08

03 Oct '08

Hi. Sorry if this discussion has already been made... Why isn't it possible to configure an smtp-server to send username(a)fedoraproject.org mails? Without it, I'm not able to properly use my fedora-mailadress - at least at the moment. Or are there any workarounds for people without an own mail-server? -Thomas

3 3

metalinks for F10 download pages
by Matt Domsch 03 Oct '08

03 Oct '08

mirrors.fp.o now supports metalinks [1]. metalinks are XML documents that act like a yum mirrorlist, but with more detail, allowing client download tools to more easily select a mirror that will be fastest for them. Just as you have the BitTorrent, Direct Download, and Jigdo columns now, I'd like to consider adding a Metalink column. It's easy to get the metalink URL for a given download.fp.o ISO URL. http://download.fedoraproject.org/pub/fedora/linux/releases/test/10-Beta/Fe… becomes http://mirrors.fedoraproject.org/metalink?path=pub/fedora/linux/releases/te… (e.g. s:download:mirrors:, s:pub(.*):metalink?path=pub$1: ) Let me know if you have any questions. If you wanted to add them to the 10-Beta download page now, that'd be fine with me too. Thanks, Matt Fedora Mirror Wrangler. [1] http://metalinker.org/ -- Matt Domsch Linux Technology Strategist, Dell Office of the CTO linux.dell.com & www.dell.com/linux

3 3

fedora hosted
by Mike McGrath 02 Oct '08

02 Oct '08

Hey guys, there's a bunch of outstanding fedorahosted requests right now: https://fedorahosted.org/fedora-infrastructure/report/1?sort=component&asc=1 Can someone go through and take care of these? -Mike

3 3

Puppet training
by Mike McGrath 02 Oct '08

02 Oct '08

Hey guys, i think I'd like to hold puppet training next week on Wed sometime. Which would work best for you guys: 1:00 pm Chicago time 4:00 pm Chicago time. The live training will be identical to this training: http://mmcgrath.fedorapeople.org/puppet/ But will allow for Q and A. The training at that links includes an ogg and takes about a half hour to complete at full speed though if you're new to puppet its worth it to stop in the middle and review some of the topics. If you have any questions or comments about it please let me know. The slideshow is made with openoffice and I made the ogg with audacity. My throat hurts now so I'm going to get some tea :) -Mike

12 20

Meeting Log - 2008-10-02
by Ricky Zhou 02 Oct '08

02 Oct '08

20:01 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Who's here? 20:01 * ricky 20:01 * mmcgrath is 20:01 * SmootherFrOgZ here 20:02 < mdomsch> yo 20:02 < G> I'm here 20:02 * nirik is sitting in the cheap seats in the back. 20:02 * mmcgrath waves at nirik 20:02 -!- ianweller_afk is now known as ianweller 20:02 * nirik waves back 20:02 < mmcgrath> Ok, lets get started 20:02 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Tickets 20:02 < mmcgrath> .ticket 753 20:03 < zodbot> mmcgrath: #753 (Mini-freeze for Fedora 10 beta) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/753 20:03 < mmcgrath> err 20:03 < mmcgrath> .tiny https://fedorahosted.org/fedora-infrastructure/query?status=new&status=assi… 20:03 < zodbot> mmcgrath: http://tinyurl.com/47e37y 20:03 < mmcgrath> f13: were you happy with the mini freeze? 20:03 < mmcgrath> this is the first time we had one, what did you all think? 20:03 < G> mmcgrath: fine 20:03 -!- MadBus [n=scott@fedora/vwbusguy] has left #Fedora-meeting ["Leaving"] 20:04 < mmcgrath> I think it went well 20:04 -!- jbasse [n=jbasse(a)104.226-241-81.adsl-static.isp.belgacom.be] has quit 20:04 < mmcgrath> f13 might be doing other things. 20:04 < mmcgrath> I'll close that ticket. 20:04 < mmcgrath> .ticket 395 20:04 < zodbot> mmcgrath: #395 (Audio Streaming of Fedora Board Conference Calls) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/395 20:04 * ianweller rolls in 20:04 < mmcgrath> jcollie: anything else there? Should I leave this as a meeting item or remove the meeting tag from it? 20:06 * mmcgrath will skip it for now 20:06 < mmcgrath> .ticket 446 20:06 < zodbot> mmcgrath: #446 (Possibility to add external links on spins page) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/446 20:06 < mmcgrath> dgilmore: ping ^^^ 20:07 -!- Cheguevara [n=Chegueva(a)host86-132-195-153.range86-132.btcentralplus.com] has quit Read error: 104 (Connection reset by peer) 20:07 * mmcgrath suspects lots of people are busy today :-/ 20:07 < mmcgrath> Ok, we'll skip that one and 740 20:07 < dgilmore> mmcgrath: forgot to close 20:07 < mmcgrath> dgilmore: yo! 20:08 < mmcgrath> dgilmore: k 20:08 < mmcgrath> and 20:08 < mmcgrath> .ticket 740 20:08 < zodbot> mmcgrath: #740 (Loaning out system time to OLPC participants) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/740 20:08 < mmcgrath> we talked to these guys at the last meeting, was there any resolution there? 20:08 -!- themayor [n=jack(a)dsl254-115-149.nyc1.dsl.speakeasy.net] has quit 20:09 < mdomsch> gregdek notified the 100 or so recepients of the systems to test 20:09 < G> we should help them build an OLPC Buildsys on XOs :P 20:09 < mmcgrath> G: using distcc? :) 20:09 < G> mmcgrath: exactly :) 20:09 * G runs away before getting murdered 20:10 -!- Cheguevara [n=Chegueva(a)host86-132-195-153.range86-132.btcentralplus.com] has joined #fedora-meeting 20:10 < mmcgrath> heh 20:10 * mmcgrath will assume nothing further has happened there. 20:10 < mmcgrath> so thats the end of the tickets 20:10 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- oVirt 20:11 < mmcgrath> So i continue to punch away at oVirt for this and that. 20:11 < mmcgrath> SmootherFrOgZ: would you be interested in partnering with me on this? 20:11 < SmootherFrOgZ> sure 20:12 < mmcgrath> SmootherFrOgZ: so after this meeting ping me and we'll work out a game plan. I'll make sure you get access to our xen6 box where I'm doing the testing and stuff. 20:12 -!- Prakhar [n=prakhar(a)122.163.71.31] has quit "Open source>open minds>open coolness!!" 20:12 < SmootherFrOgZ> k 20:12 < mmcgrath> There's really not been much progress there this last week though, nothing to report from last week anyway. 20:13 < mmcgrath> So with that I'll move on 20:13 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Zabbix 20:13 < mmcgrath> So I've been poking around at zabbix, making some custom queries, maps, etc. Generally just getting to know it better. 20:13 < mmcgrath> G: do you want to give a general roadmap for what is next? 20:14 < G> Okay here goes... 20:14 -!- spot_ [i=spot(a)beer.tclug.org] has joined #fedora-meeting 20:14 -!- spot [i=spot@redhat/spot] has quit Remote closed the connection 20:15 < G> We need to upgrade to 1.6.0 which is a givin, I don't see any problem here, however, I've noticed a couple of evil little bugs that'll spoil our game plan big time 20:15 -!- spot_ is now known as spot 20:15 < G> I hope to spend next week looking at it all so we don't run into brick walls 20:15 < mmcgrath> bugs in the upgrade or in 1.6.0 itself? 20:15 < mmcgrath> <nod> brick walls are bad. 20:15 < G> mmcgrath: bugs in 1.5.4 that I'm not sure if they have been fixed in 1.6.0 20:16 -!- giallu [n=giallu@fedora/giallu] has joined #fedora-meeting 20:16 -!- viking-ice [n=johannbg(a)157-157-119-239.dsl.dynamic.simnet.is] has joined #fedora-meeting 20:16 < mmcgrath> <nod> 20:16 < mmcgrath> and these are bugs that we couldn't live with or just would be good to have fixed before we go live? 20:16 < G> Then it's just a matter of making sure our templating is correct, and removing zabbix from all the manifests files and putting in global (one reason why I had not added any lately) 20:17 < G> mmcgrath: there is one wrt users, that we'd have to fix before hand 20:17 < mmcgrath> FWIW, I added the zabbix agent to global just this afternoon :) 20:17 < G> mmcgrath: oh I didn't notice that, I was going to sort everything else out before hand to make it easier 20:17 < mmcgrath> cool 20:18 < mmcgrath> are we going to do a clean wipe and fresh install or an upgrade? 20:18 < ricky> Do we want zabbix running on backup1 on the next puppet run? 20:18 < G> mmcgrath: that was one of the reasons that I hadn't done anything like that 20:19 < G> I just _don't_ know whats going to have to happen with the upgrade, really it depends if I can reproduce some of these bugs on a clean install 20:19 < mmcgrath> ricky: not sure, its something to examine. 20:19 < ricky> OK 20:19 < G> I can't do anything like that until Monday though 20:19 < mmcgrath> G: well I guess worst case we can start over. Thats the nice thing of considering it a beta :) 20:19 < G> mmcgrath: exactly 20:19 < mmcgrath> yeah, no rush on this. I'm thinking more along "sometime before 2009" 20:20 < mmcgrath> but I know it'll take a while once we get going anyway, though really its been quite easy to add things. 20:20 < G> I guess I should add the following: 20:21 < G> My avalibility is going to be all up the creek for the end of this month and start of November, so it's likely to be completed later rather than sooner 20:21 < mmcgrath> <nod> thanks for the heads-up :) 20:21 < mmcgrath> anyone have anything else to add on monitoring? 20:21 < G> (I can't say why yet) 20:22 < mmcgrath> k 20:22 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Ibiblio 20:22 < mmcgrath> So our ibiblio site is up and online. torrent is running from there as is proxy4 (soon) 20:22 < mmcgrath> So woot! 20:22 < G> nice 20:22 < mmcgrath> and make sure you thank ibiblio folk if you see them around. 20:23 -!- mmcgrath changed the topic of #fedora-meeting to: Infrasturcture -- Open Floor 20:23 < mmcgrath> So thats really all I had. 20:23 < mmcgrath> anyone have anything they'd like to discuss? 20:23 < G> Errr yeah, watch out for massive changes to the elections app 20:23 < ricky> Might be a good idea to add the ibiblio contact info in the private repo 20:23 < mmcgrath> hahaha 20:23 < mmcgrath> whats going on there? 20:23 < mmcgrath> ricky: I'll make sure it gets in there soon, putting it on my TODO now 20:23 < G> mmcgrath: FEATURES! 20:24 < ricky> Thanks 20:24 < mmcgrath> features or featurecreep? 20:24 < G> spoleeba reminded me of some useful features that I'd been meaning to implement for a while 20:24 < mdomsch> he's good at that 20:24 < G> so it's likely I'll be doing quite a few changes 20:24 < nirik> any progress on ticket 651? or is that on hold after the intrusion fun? 20:25 < ricky> .ticket 651 20:25 < mmcgrath> .ticket 651 20:25 < zodbot> ricky: #651 (system account for tier2 ppc32/x86_64 hosting for cvsextras) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/651 20:25 < zodbot> mmcgrath: #651 (system account for tier2 ppc32/x86_64 hosting for cvsextras) - Fedora Infrastructure - Trac - https://fedorahosted.org/fedora-infrastructure/ticket/651 20:25 < G> nirik: denied, there is no group cvsextras :) 20:25 -!- viking-ice [n=johannbg(a)157-157-119-239.dsl.dynamic.simnet.is] has quit Remote closed the connection 20:25 < mmcgrath> I want to say thats ready but I don't know the specifics 20:25 < G> (he lies!) 20:25 < nirik> G: there was when that was filed. ;) 20:25 < G> nirik: sure sure ;) 20:25 < mmcgrath> ricky: can you take a look on fas1 and see if that status is ready? 20:25 < ricky> nirik: loupgaroublond and I are working on fasClient a bit, so we'll try to take that into consideration 20:26 < mmcgrath> its basically the "use fasClient but don't send passwords" 20:26 < mmcgrath> ricky: I think loupgaroublond may have already finished that part. 20:26 < G> so yeah, wrt elections app there may be some change requests during the freeze 20:26 < nirik> cool. Happy to test from here... 20:26 < mmcgrath> but if you guys are re-doing everything perhaps we should wait. 20:26 < nirik> just let me know when you want me to. 20:26 < G> Which reminds me, when is the final release Freeze? 20:26 < mmcgrath> ricky: do you and loupgaroublond have a time frame on when the new fasClient implementation might be ready? 20:27 < mmcgrath> G: the final freeze starts on Nov 11 20:27 < ricky> Hopefully in the next week or two 20:28 < ricky> I think we have the basic data-fetching thought out, but we just need to make sure that all the cases/proper behavior are covered in the end 20:28 < G> mmcgrath: brilliant 20:28 < mmcgrath> looks like we'll have another pre-freeze on october 21st 20:28 -!- Cheguevara_ [n=Chegueva(a)host86-132-195-153.range86-132.btcentralplus.com] has joined #fedora-meeting 20:28 < mmcgrath> ricky: solid 20:28 -!- Cheguevara [n=Chegueva(a)host86-132-195-153.range86-132.btcentralplus.com] has quit Read error: 104 (Connection reset by peer) 20:28 < mmcgrath> Anyone have anything else they'd like to discuss? 20:28 < G> mmcgrath: I don't think so 20:29 < mmcgrath> sweet, if there's nothing else I'll close the meeting in 30 20:29 < mmcgrath> 10 20:29 < mmcgrath> 5 20:29 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Meeting end

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

infrastructure October 2008