infrastructure April 2008

infrastructure@lists.fedoraproject.org

52 participants
89 discussions

Introduction
by German G. Ramirez 13 Apr '08

13 Apr '08

Dear fedora-infrastructure-list, Permit me to introduce myself.My name is German I'm a 26 y.o Mexican living in Kiev,Ukraine.I own a bachelor degree in computer science and have had work experience as system administrator and programming.I'm currently running fedora 8 on my laptop.I do not have experience using Python but I'm willing to learn it.Since fedora is my new hobby I would like to contribute to the fedora project and work with people sharing the same interests just for the fun of doing it. Looking forward to work with you all. Regards, German

1 0

app servers updated
by Toshio Kuratomi 12 Apr '08

12 Apr '08

I've built and added to the fedora-infrastructure repo the following packages: * fas-clients -- Contains a backported patch so we don't see deprecation warnings when run with the new python-fedora. * fas -- Contains one backported patch that we're currently running on fas1 and fas2. * python-sqlalchemy -- Major version upgrade. from 0.3 => 0.4. AFAIK this update only affects smolt and packagedb, both of which have been ported and tested with this release. * fedora-packagedb -- Major changes to support SQLAlchemy 0.4 and to re-optimize some code paths for FAS2. This release should be much faster for certain operations like viewing the firefox page or mass branching. * python-fedora -- Major changes. removed all FAS1 code. Moved BaseClient and deprecated the old location. Updated json.py for SQLAlchemy-0.4. Added a kludge for bz_email handling. Threaded jsonfasvisit so it contacts the fas server less often. fas-clients, python-sqlalchemy, fedora-packagedb, and python-fedora have been installed on app2, app4, and app5.vpn. These are our TurboGears app servers. I haven't installed them on fas1 & fas2 or any of the other machines yet. We can do that if we want or wait until after the change freeze. *If there's any problems with the new packagedb or other TurboGears apps please contact me so I can look at fixing it or rolling back these changes before the change freeze. (abadger1999 on freenode)* The next version of python-fedora will likely have some incompatible changes to the interface between BaseClient and the Server. Those changes will make error handling between the Server and Clients better and standardize some of the things that we currently do in an ad hoc fashion but will require modifications of all the servers that talk to clients in order to work. I'll let people know when I have a new version of BaseClient ready to test against. The specification from the server side view is here:: http://bzr.fedorahosted.org/bzr/python-fedora-python-fedora-devel?cmd=conte… -Toshio

1 0

Fedora CA Project
by Dennis Gilmore 11 Apr '08

11 Apr '08

We have come to the realisation that this has to be done sooner rather than later. So i'm putting out a call for help and for feedback. We need to revamp the CA infrastructure used in Fedora. This is where Id like to see us go. Publish a Certificate Revocation list so that all apps can check for revoked certs Have users able to revoke their own cert Have user certs be revoked when they request a new cert Have admins able to create/revoke certs Their are 2 types of certificates currently handled by 2 CA's I really want to use a single CA for all: Type 1) user certs. used for plague/koji/cvs upload access. there is work underway to use these for other fedora web based apps also. Type 2) Builders, kojira, internal service authentication. Products to be evaluated: http://pki.fedoraproject.org/wiki/PKI_Main_Page https://www.openca.org/ http://ejbca.sourceforge.net/ Something custom FAS will need modification to work with the new framework. I also want to allow fedora-packager-setup to grab the cert directly rather than having the user manually do it. probably with a flag for when to get a new cert. All users will need to get new user certs when we make the change. as well as koji hub, all builders, koji garbage collection, bodhi, It would also be a good time to deploy ssl auth for other apps. We have a ticket https://fedorahosted.org/fedora-infrastructure/ticket/466 Please make suggestions for other apps we could use, also ideas for making the workflow better. So this is a brief overview of whats needed. Im going to open the floor for a week for open discussion on how we should best do this. Dennis

8 12

Meeting Log - 2008-04-10
by Ricky Zhou 11 Apr '08

11 Apr '08

15:00 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Who's here? 15:01 * yingbull is. 15:01 < mmcgrath> Alrighty everyone, who's around? 15:01 * warren here 15:01 * skvidal is here 15:01 * notting is here 15:01 < mmcgrath> jcollie: you around? 15:02 < jcollie> yup 15:02 < ivazquez> Pong. 15:02 < mmcgrath> cool, lets get started then 15:02 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Tickets 15:02 < mmcgrath> .tiny https://fedorahosted.org/fedora-infrastructure/query?status=new&status=assi… 15:02 < zodbot> mmcgrath: http://tinyurl.com/2hyyz6 15:02 * MostafaDaneshvar is there 15:02 < mmcgrath> First ticket: 15:02 < mmcgrath> .ticket 395 15:02 < zodbot> mmcgrath: #395 (Audio Streaming of Fedora Board Conference Calls) - Fedora Infrastructure - Trac - https://fedorahosted.org/projects/fedora-infrastructure/ticket/395 15:02 < mmcgrath> jcollie: whats the latest? 15:02 < warren> mmcgrath: is agenda set in stone? I want to add s/cvsextras/packager/ after F9 release and next convenient scheduled outage. 15:03 < mmcgrath> warren: not at all, we pretty much go over tickets now, then a few other items then open the floor. 15:03 < mmcgrath> warren: Remind me if I forget at the end. 15:03 < jcollie> not much has happened... someone horked up a bunch of my virtual hosts at work so i've been busy rebuilding them 15:03 < mmcgrath> jcollie: virtual hosts at $DAYJOB or something related to fedora? 15:03 -!- Wakko666 [n=blentz(a)office.cardomain.com] has joined #fedora-meeting 15:03 < jcollie> $DAYJOB 15:03 -!- wolfy [n=lonewolf@fedora/wolfy] has left #fedora-meeting ["When you breathe, you inspire. When you do not breathe, you expire."] 15:04 < mmcgrath> just checking ;-) 15:04 < mmcgrath> jcollie: k, so nothing new there. We'll move on. 15:04 < mmcgrath> .ticket 398 15:04 < zodbot> mmcgrath: #398 (elfutils `monotone' (mtn) error) - Fedora Infrastructure - Trac - https://fedorahosted.org/projects/fedora-infrastructure/ticket/398 15:04 * mmcgrath sees if rmcgrath is around. 15:04 < jcollie> i think that those of us that are interested in working on the streaming should set up a separate meeting to divvy up some tasks 15:05 < mmcgrath> jcollie: not a bad idea. It sounds like you've done most of the work already. Its just a matter of getting that last 10% done (and getting the fas integration for the non streamins stuff) 15:05 < jcollie> yep 15:06 < mmcgrath> alrighty, we'll talk about 398 later if rmcgrath becomes available. 15:06 < mmcgrath> My understanding is that we can do it though. 15:06 < mmcgrath> .tickety 446 15:06 < mmcgrath> .ticket 446 15:06 < zodbot> mmcgrath: #446 (Possibility to add external links on spins page) - Fedora Infrastructure - Trac - https://fedorahosted.org/projects/fedora-infrastructure/ticket/446 15:06 < mmcgrath> doath 15:06 * dgilmore is here 15:07 < mmcgrath> dgilmore: I think last time we talked about you getting this up on the wiki, any news on that? 15:07 < dgilmore> mmcgrath: i started and got sidetracked 15:07 < mmcgrath> dgilmore: ahh, you're still on it though? 15:07 < dgilmore> mmcgrath: will be done by the end of the weekend 15:07 < dgilmore> yep 15:08 < mmcgrath> AFAIK we just have the one spin so far. 15:08 < mmcgrath> dgilmore: sweet, thanks! 15:08 < mmcgrath> Ok, so thats really the end of it on the tickets side. 15:08 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- New Wiki 15:08 < mmcgrath> Just a bit of a roundup on the wiki. I've been working on th emigration script and docbook exporter script. 15:08 < mmcgrath> The people listed on http://fedoraproject.org/wiki/Infrastructure/WikiMigration all met yesterday. 15:09 < mmcgrath> I'm going to schedule a few more meetings but a few bumps aside it sounds like we are a GO. 15:09 < mmcgrath> dgilmore: are you on the FESCo ? 15:09 < dgilmore> mmcgrath: i am 15:09 < mmcgrath> dgilmore: at the next meeting would you find a delegate to sign up to be the wiki liason for those sections of the wiki? 15:10 < mmcgrath> unless you want to do it :) 15:10 < dgilmore> mmcgrath: sure 15:10 < mmcgrath> Thanks. 15:10 < dgilmore> mmcgrath: i think jwb volunteered 15:10 * jwb wakes up 15:10 < jwb> huh? 15:10 * dgilmore notes he will largely be away next week 15:10 < mmcgrath> so https://publictest1.fedoraproject.org/wiki/FedoraMain is still up and ready 15:10 < dgilmore> jwb: wiki migration? 15:10 < mmcgrath> dgilmore: ohhh yeah. you're right. he's on there. 15:11 < jwb> dgilmore, mmcgrath: oh, yeah. i put my name there 15:11 < mmcgrath> dgilmore: I'll be away next week as well. 15:11 < mmcgrath> jwb: sorry I missed it there. 15:11 < jwb> i have no idea what that means 15:11 < jwb> do i have to do something? 15:11 < jwb> it just said "contact point" 15:11 < dgilmore> jwb: you get to be whiped 15:11 < mmcgrath> jwb: not yet, we'll be scheduling a meeing again the week after next. In the meantime just go through https://publictest1.fedoraproject.org/wiki/ and look for... doom. 15:11 < dgilmore> jwb: no Just be a point person. 15:12 < jwb> ok great 15:12 < jwb> either way, count me in 15:12 < mmcgrath> cool. 15:12 < mmcgrath> So thats really the latest on the wiki. 15:12 < mmcgrath> Anyone have any questions or concerns? 15:13 < mmcgrath> allrighty, moving on 15:13 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Next week 15:13 < mmcgrath> I'll be gone again next week for training. Its the last traing session I actually have scheduled. 15:13 < mmcgrath> So during the days I'll be unavailable but I'll likely be around most evenings. 15:14 < mmcgrath> Next topic 15:14 < dgilmore> mmcgrath: ill be in Brazil 15:14 * skvidal will be here 15:14 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- nfs1 15:14 < skvidal> people can call/page me 15:14 < skvidal> ugh 15:14 < skvidal> nfs1 15:14 < mmcgrath> NFS1 got built yesterday in a hurry. 15:14 < mmcgrath> while working on xen2 the nfslock issues showed up again then the whole box took a panic. 15:15 < mmcgrath> we'd been serving nfs directly from xen2 so when it goes down the buildsystem goes down. 15:15 < mmcgrath> Anyway, we brought nfs1 up as a RHEL4 box to try to fix the nfs issues, they might have fixed nfs issues but brought on ext3 issues since we're using a 10T ext3 filesystem... 15:15 < mmcgrath> RHEl4 couldn't handle it and started to ioerror. 15:15 < mmcgrath> So I kicked a RHEL5 box and its now nfs1. 15:16 < mmcgrath> Its up, its running. 15:16 < mmcgrath> so far no nfslock issues, if we do get one though we're going to contact steved while its in the bad state and give him a shell to look around. 15:16 < dgilmore> mmcgrath: lets get steved a shell on xen2 and let him monitor nfs1 15:16 < mmcgrath> Really nothing new here except that /mnt/koji is no longer available on xen2, its now on nfs1. 15:17 < mmcgrath> dgilmore: you never know, perhaps magically moving to nfs1 fixed our problem :) 15:17 < mmcgrath> we'll see. 15:17 * skvidal likes magic 15:17 < dgilmore> mmcgrath: we can hope 15:17 < dgilmore> skvidal: voodoo magic is best 15:17 < mmcgrath> heheh 15:17 * skvidal makes a doll of dgilmore 15:17 < mmcgrath> Ok, next item. 15:17 < dgilmore> skvidal: need some hair and fingernails? 15:18 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- s/cvsextras/packagers/ 15:18 < mmcgrath> warren: ping? 15:18 < warren> hi 15:18 < mmcgrath> warren: so we're just going to rename cvsextras to packagers. 15:18 < mmcgrath> which involves 15:18 < warren> So f13 suggested we rename the group after F9 release, at the first scheduled outage. 15:18 < warren> mmcgrath: packager, singular 15:18 < mmcgrath> 1) updating the db. 15:19 < mmcgrath> 2) updating the scripts (and upload script) 15:19 < mmcgrath> 3) updating the documentation 15:19 < mmcgrath> 4) testing? 15:19 < mmcgrath> I really don't think this will be that huge of a deal. 15:19 < mmcgrath> warren: I'll try to get it done sometime the week after F9 ships. Would you mind opening a ticket and bugging me about it that week? 15:19 < warren> mmcgrath: ok 15:20 < warren> mmcgrath: are there any other group names we should rename as well? 15:20 < f13> warren: lets keep the name that will make sense in the future given my newmaintianercontainment proposal 15:20 < warren> as long as we have an outage 15:20 < f13> warren: since I hope to get an intern to work on that this summer 15:20 < warren> f13: is this the one from 2 months ago or so? 15:20 < mmcgrath> warren: yeah we'll schedule an outage. 15:20 -!- Milanito [n=Matt(a)32.12.70-86.rev.gaoland.net] has quit Remote closed the connection 15:20 < f13> warren: http://fedoraproject.org/wiki/JesseKeating/NewMaintainerContainment 15:21 < mmcgrath> I'm not sure about other group names. 15:21 < warren> Let's ask around 15:21 < warren> there might be other names people want to do 15:21 < warren> f13: ok seems we need to figure out a bigger picture plan then 15:22 < warren> mmcgrath: will get back to you 15:22 < mmcgrath> warren: works for me, lets know well in advance of the outage know because renaming cvsextras I have a pretty good idea of what needs to be done. 15:22 < mmcgrath> but for another group I might not know without more research. 15:22 < mmcgrath> k 15:22 < mmcgrath> warren: when its all ready just create a ticket and bug me. I think we could have it done in an hour or two some night after F9 ships. 15:22 < f13> warren: well, we don't necessarly have to wait to rename 'cvsextras' to /somethihng/ 15:23 < f13> warren: 'fedorapackager' or whatever seems fine, we can pick a name that implies /greater/ rights later 15:23 < warren> f13: I do agree to rename it within the larger plan 15:23 < f13> or less rights. 15:23 < warren> f13: I'm still in favor of numbers instead of group names for that 15:23 < warren> but we're getting off topic 15:23 < mmcgrath> <nod> you guys figure that out :) 15:24 < mmcgrath> anyone have any questions about that? We've got plenty of time to discuss it. 15:24 < mmcgrath> if not I'll move on. 15:24 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- app server upgrades. 15:24 < mmcgrath> abadger1999: ping? 15:25 < mmcgrath> I believe toshio wants to upgrade sqlalchemy and python-fedora before the freeze. I think we're on schedule to do that but I'll have to get ahold of him to find out for sure. 15:25 < mmcgrath> which brings me to the next topic 15:25 < abadger1999> here. 15:25 < mmcgrath> oh 15:25 < mmcgrath> there you are :) 15:25 < abadger1999> Today. 15:26 < mmcgrath> abadger1999: want to talk about all of that right quick? 15:26 < abadger1999> Want to do it today/tonight. 15:26 < abadger1999> It would be to both app servers. 15:26 < mmcgrath> abadger1999: "both" ? 15:26 < mmcgrath> which? 15:26 < abadger1999> app4 and app5 15:26 < mmcgrath> app2 is also running tg apps 15:26 < abadger1999> Oh. I was not aware. 15:27 < dgilmore> mmcgrath: i want the new python-fedora 15:27 < abadger1999> then we wantto upgrade that as well. 15:27 < abadger1999> We can issue a new python-fedora for everything if we want. 15:27 < mmcgrath> abadger1999: your call I'm not sure what we'd be comitting to with the new package. 15:27 < abadger1999> Or we can update that when we do the other updates from RHEL, etc. 15:27 < mmcgrath> what are the risks / benefits? 15:27 < mmcgrath> and will this break transifex? 15:28 < abadger1999> All the FAS1 stuff has been removed. Only FAS2 stuff is in it. 15:28 < abadger1999> there's been some rearrangement of modules but the old imports should work with a deprecation warning. 15:28 < mmcgrath> k 15:29 < mmcgrath> sounds pretty low risk, and we can always revert if required. 15:29 < abadger1999> <nod> 15:29 < mmcgrath> abadger1999: I'm fine for any time this week on that, if you want me around just ping me or call me. 15:29 < mmcgrath> abadger1999: anything else on that? 15:30 < abadger1999> Nope that's it. 15:30 < mmcgrath> cool. 15:30 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- app5 15:30 < mmcgrath> So yeah, I don't know WTF is going on but app5 is rebooting... a LOT 15:30 < mmcgrath> this comes after the move from x86_64 to i686 15:30 -!- Milanito [n=Matt(a)32.12.70-86.rev.gaoland.net] has joined #fedora-meeting 15:30 < mmcgrath> we're talking between 20-40 times a DAY. 15:30 < mmcgrath> very very strange. 15:30 < mmcgrath> I've got a stack trace, I'm going to try to take it to the virt guys soon. 15:31 < mmcgrath> side note though, aside from that, the i686 move (also done on app2) has proved quite successful memory usage (and therefor swap) is way down. 15:31 < mmcgrath> its been a big win for us. 15:31 < mmcgrath> anyone have any questions / comments? 15:31 -!- viking-ice [n=johannbg(a)dsl-149-97-225.hive.is] has joined #fedora-meeting 15:31 < mmcgrath> Alrighty, then I'll move on to the last topic of the day before opening the floor. 15:32 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Change Freeze! 15:32 < mmcgrath> This one's for everyone 15:32 < mmcgrath> REMEMBER 15:32 < mmcgrath> DO NOT MAKE ANY CHANGES WITHOUT GETTING APPROVAL ON THE LIST FIRST! 15:32 < mmcgrath> and remember, you'll be asked the question "Why can't this wait until after F9 gets released" 15:33 < mmcgrath> I'll be gone much of next week which makes the freeze perfect. 15:33 < warren> under pain of 15:33 * mmcgrath looks up exact date again 15:33 < mmcgrath> warren: under pain of CURSE! 15:33 < skvidal> under pain of pain 15:33 < warren> I can advise against pain. 15:34 < mmcgrath> There we go, so the freeze starts the 15th. The release is on the 29th. The freeze is lifted on the 30th. 15:34 < yingbull> pain's scary. 15:34 < mmcgrath> Obvious exceptions include outages and things already scheduled for the release (mm type stuff, the website, etc) 15:34 < mmcgrath> other then that. you must get approval on the list before making changes. If you're working on tickets and people get cranky, just let them know we're under a freeze. 15:34 < skvidal> mmcgrath: it's okay to bring up new boxes unrelated to the release? 15:34 < ivazquez> Should we hold it until the Monday following? 15:35 < skvidal> mmcgrath: ie: if I get all the statics, etc from BU 15:35 < mmcgrath> skvidal: take it to the list! But yeah, that shouldn't be a problem. 15:35 < skvidal> mmcgrath: will do 15:35 < mmcgrath> ivazquez: I thought about that, in the past though, for the most part, things settled down from the release the day after. 15:36 < ivazquez> Okay. 15:36 < mmcgrath> anyone have any questions / comments about that? 15:36 < mmcgrath> alrighty, then we'll open the floor 15:36 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Open Floor 15:36 * dgilmore has nothing right now 15:36 < mmcgrath> Anyone have anything they'd like to discuss? A release is coming, lots of stuff has been going on. 15:37 < notting> how are we ensuring the ability to get the bits to our tier0 mirrors? 15:37 < mmcgrath> notting: same as we had been, AFAIK our mirror system (and I2) is functional again. 15:37 < notting> that's not what mdomsch's mail implied 15:37 < mmcgrath> notting: which email and when? 15:38 < notting> to mirror-list 15:38 < dgilmore> notting: no, it implied a work around was setup 15:38 < dgilmore> mmcgrath: a couple of hours ago 15:38 < dgilmore> mmcgrath: 2 boxes were setup for Tier 0 I2 masters 15:39 < mmcgrath> <nod> that was my understanding. We don't have the permanent solution in place but a working one. 15:39 < mmcgrath> we'll probably want to make sure the working one stays as is until after the 30th. 15:39 < dgilmore> mmcgrath: they have static I2 routes to dl.fedora.redhat.com boxes 15:39 < mmcgrath> notting: I'll follow up with mdomsch to make sure whats going on there. 15:39 < mmcgrath> .any mdomsch 15:39 < zodbot> mmcgrath: mdomsch was last seen in #fedora-meeting 1 hour, 9 minutes, and 24 seconds ago: *** mdomsch has quit IRC (Remote closed the connection) 15:39 -!- Milanito [n=Matt(a)32.12.70-86.rev.gaoland.net] has quit Remote closed the connection 15:40 < mmcgrath> Alrighty, anyone have anything else to discuss? 15:40 < dgilmore> nope 15:40 < mmcgrath> k, we'll close the meeting in 30 15:41 < mmcgrath> 15 15:41 < mmcgrath> 5 15:41 < notting> dgilmore: so, the static route & download1 are bandwidth-unclogged enough to sync OK? 15:41 < dgilmore> Thanks mmcgrath :) 15:41 < notting> dgilmore: with no one using it, are we sure that download1 is synced? :) 15:41 < dgilmore> notting: i believe so. 15:42 < mmcgrath> notting: we'll have to talk to them to find out for sure. 15:42 * mmcgrath just sent an email to mdomsch to verify. 15:42 < notting> maybe i'm paranoid 15:42 < mmcgrath> notting: you're once bitten. 15:42 < mmcgrath> nothing wrong with that. The fact is the whole I2 connection thing showed a major problem we have. and the fact that galgoci is the _only_ person at Red hat with access and no how to fix it is much less comforting. 15:43 < dgilmore> notting: i believe download1 is down 15:43 < mmcgrath> even his backup told us that he was really the only guy we'd want doing that. 15:44 -!- k0k [n=k0k@fedora/k0k] has quit Connection timed out 15:44 < mmcgrath> notting: we'll just have to wait and see if what we think mdomsch said is actually the way things are. That mirror has been down since before Feb 20th so its hard to say the state of things (considering they only just recently got back up and running) 15:44 < mmcgrath> Anywho, not much we can do but check with the mirrors list and mdomsch and make sure everyone's happy with where things are for the release. 15:44 < mmcgrath> Anyone have anything else? If not we'll close the meeting in 30 15:45 -!- Milanito [n=Matt(a)32.12.70-86.rev.gaoland.net] has joined #fedora-meeting 15:45 < mmcgrath> 10 15:45 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Meeting Closed

1 0

Re: [Fedora-legal-list] Re: Possible new CLA
by Paul W. Frields 10 Apr '08

10 Apr '08

On Wed, 2008-04-09 at 11:58 -0400, Tom "spot" Callaway wrote: > On Wed, 2008-04-09 at 15:43 +0000, Paul W. Frields wrote: > > > I think it's spot. He'll see this and confirm or deny as needed. :-) > > CLAs come into my inbox. I hadn't been keeping them, just sanity > checking them. If I understand what Ricky meant by "log table" correctly, we have a list of dates that we can tie to account names. If someone could confirm this who understands the bits involved, that would be great. -- Paul W. Frields http://paul.frields.org/ gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717 http://redhat.com/ - - - - http://pfrields.fedorapeople.org/ irc.freenode.net: stickster @ #fedora-docs, #fedora-devel, #fredlug

3 2

Possible new CLA
by Paul W. Frields 09 Apr '08

09 Apr '08

There is a draft of a new CLA being circulated at FreeIPA: https://www.redhat.com/archives/freeipa-devel/2008-April/msg00052.html http://www.freeipa.org/wiki/images/2/2b/GenericCLA.pdf I've asked Legal for a summary of what's changed and why, and an answer to whether all Fedora contributors would need to re-execute this agreement. (I'm betting the answer to the last question is "yes," but that's just an educated guess.) I believe the changes to be minor, mainly to eliminate loopholes and to make this process less onerous in the future, but I'll wait for more information. Ricky Zhou tells me that our CLA signing information is kept in the form of a log table, not a simple date field. AIUI then, we have the ability to find out what CLAs have been executed over the life of an account. If I'm correct about that, I want to say first, thanks for a really good design. :-) Second, since a new CLA is a possibility -- I just want to confirm that we are not *overwriting* old CLA data, i.e. we would maintain a history that user "johndoe" signed a CLA on 2007-10-15 and then again on 2008-04-28. -- Paul W. Frields http://paul.frields.org/ gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717 http://redhat.com/ - - - - http://pfrields.fedorapeople.org/ irc.freenode.net: stickster @ #fedora-docs, #fedora-devel, #fredlug

2 2

download*.fedora.redhat.com rsyncd change request
by Matt Domsch 07 Apr '08

07 Apr '08

On download*.fedora.redhat.com, please make the following changes to the rsyncd.conf file. 1) add a motd file, with content such as: Fedora Master Mirror Servers See http://fedoraproject.org/wiki/Infrastructure/Mirroring for instructions. Modules for Fedora Core and Extras have been removed, as this content is no longer updated. See the instructions above for how to mirror current content. 2) remove fedora-linux-core* and fedora-linux-extras rsync modules. In doing so, we will hopefully be put back in contact with the ~80 mirrors who never made the transition from core/ to releases/. Thanks, Matt Domsch Fedora Mirror Wrangler -- Matt Domsch Linux Technology Strategist, Dell Office of the CTO linux.dell.com & www.dell.com/linux

1 0

fedora-infrastructure.git
by Roland McGrath 07 Apr '08

07 Apr '08

What is the relationship between git://git.fedorahosted.org/git/fedora-infrastructure.git and the configs/system scripts in puppet1:/cvs/puppet? If I want to commit a new version of a script and then get it deployed, what is the procedure? Thanks, Roland

2 1

download2.fedora.redhat.com is behind; download3 overloaded
by Matt Domsch 06 Apr '08

06 Apr '08

Since download1 DNS is pointing at download3 right now, rsync connections to that box often fail. I think it's overloaded. Recall download3 is in PHX. Download2 is running "behind" in its snap-mirroring. Or at least that's the impression one gets when rsyncing from it - after several several hours, it still hasn't finished picking up all the rawhide pushes from each day. It has been this way since at least mid last week; I don't have better statistics than this. As rawhide is now pushed from netapp in PHX, I suspect the overload of download3 is also causing bandwidth to be unavailable for snap-mirroring from the netapp in PHX to download2 in TPA. As it stands, I am extremely worried about the Fedora 9 launch scheduled for 29 April, data to flow to the mirrors starting ~24 April if all goes well. That's about 2.5 weeks from now. If the unavailability of download1 in RDU from Internet2 continues, we won't be able to get the bits to the mirrors. While we have intentionally hidden download.fedora.redhat.com from the published mirrorlists (since the Fedora 8 launch), if our mirrors don't get the bits, users _will_ come searching for them, and _will_ find them on download.fedora.redhat.com. I would hate to have a repeat of the Fedora 6 launch fiasco, where we knocked redhat.com offline for a week. What can we do, immediately, to get download1 again reachable over I2, even if it's only via static routes to fedora-archives.ibiblio.org? 9 of our 10 "Tier 1" mirrors reach us over I2. We can direct them all to fedora-archives.ibiblio.org, if we can get the content over to them in a timely manner. (fwiw, 3 of the 4 mirrors.kernel.org servers aren't on I2, but all our other Tier 1 mirrors are.) I suspect this will resolve the overload of download3, and the network bandwidth challenge (which I presume exists) which will resolve the snap-mirror to download2. Thanks, Matt Fedora Mirror Wrangler, and one who doesn't want to see the fallout of knocking redhat.com offline for a week again... -- Matt Domsch Linux Technology Strategist, Dell Office of the CTO linux.dell.com & www.dell.com/linux

1 0

Is someone rebooting app5?
by Toshio Kuratomi 04 Apr '08

04 Apr '08

Did I miss a memo or is there a problem with app5? It seems to be rebooting pretty frequently. (The Xen host is not rebooting, just this guest) /var/log/messages:: [...] Apr 3 21:27:28 app5 kernel: Linux version 2.6.18-8.el5xen Apr 3 22:23:57 app5 kernel: Linux version 2.6.18-8.el5xen Apr 3 23:04:08 app5 kernel: Linux version 2.6.18-8.el5xen Apr 3 23:22:41 app5 kernel: Linux version 2.6.18-8.el5xen Apr 4 00:50:52 app5 kernel: Linux version 2.6.18-8.el5xen Apr 4 01:40:06 app5 kernel: Linux version 2.6.18-8.el5xen Apr 4 02:11:59 app5 kernel: Linux version 2.6.18-8.el5xen Apr 4 02:56:31 app5 kernel: Linux version 2.6.18-8.el5xen Apr 4 03:17:34 app5 kernel: Linux version 2.6.18-8.el5xen Apr 4 04:31:15 app5 kernel: Linux version 2.6.18-8.el5xen Logs on xen9 Show this:: [2008-04-04 03:17:07 xend.XendDomainInfo 3444] DEBUG (XendDomainInfo:988) XendDomainInfo.handleShutdownWatch [2008-04-04 04:30:42 xend.XendDomainInfo 3444] WARNING (XendDomainInfo:923) Domain has crashed: name=app5 id=53. [2008-04-04 04:30:42 xend.XendDomainInfo 3444] DEBUG (XendDomainInfo:1566) XendDomainInfo.destroyDomain(53) Services on app5 are load balanced on app4 so we aren't losing a service when this happens. -Toshio

2 1

← Newer
1
2
3
4
5
6
7
8
9
Older →

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

infrastructure April 2008