infrastructure April 2009

infrastructure@lists.fedoraproject.org

49 participants
51 discussions

sphinx as search solution?
by Ian Weller 28 Apr '09

28 Apr '09

Here's a search engine solution[1] that doesn't run on Java and has a MediaWiki plugin[2]. Please take a look at let me know what you think with regards to using it for Fedora Infra's purposes. [1] http://www.sphinxsearch.com/ [2] https://secure.wikimedia.org/wikipedia/mediawiki/wiki/Extension:SphinxSearch -- Ian Weller <ian(a)ianweller.org> GnuPG fingerprint: E51E 0517 7A92 70A2 4226 B050 87ED 7C97 EFA8 4A36

4 3

changes post-freeze
by Matt Domsch 28 Apr '09

28 Apr '09

A few things I'd like to change starting tomorrow (post-freeze). * change the MM update-master-directory-list cronjob to start at 0 and 30 past the hour, from its current schedule of trying to start every 15 minutes. It is taking about 20 minutes on average to run, so really is only running twice an hour anyhow. * bump back the MM update-mirrorlist cronjob to start at :40 past the hour. It takes about 20 minutes to complete, and I would like the new content to land at the top of the hour. * in modules/rsync/files/rsyncd.conf.secondary1, exclude alt/stage. Mirrors shouldn't be able to sync this content. * in MM prod.cfg, exclude pub/alt/stage. Mirrors shouldn't have this content, and it's extra directory walks we don't need. * increase the number of crawlers, from 45 to 75. A full run is taking about 3 hours now, I'd like to bring this down to under 2. This only affects bapp1, whose load average is still under 1 and has plenty of free RAM and CPU it seems. Objections or comments? Thanks, Matt -- Matt Domsch Linux Technology Strategist, Dell Office of the CTO linux.dell.com & www.dell.com/linux

2 2

Preview release is out!
by Mike McGrath 28 Apr '09

28 Apr '09

Get to installing. Get to testing. Get to filling out this matrix: https://fedoraproject.org/wiki/QA:Fedora_11_Preview_Install_Test_Results -Mike

1 0

Statistics problem
by Paul W. Frields 27 Apr '09

27 Apr '09

I found that our statistics for downloading have not been properly capturing requests for Live ISO images from our download links on the fp.o/get-fedora page. I'll be updating the numbers on the wiki's [[Statistics]] page in the next day or two. I did two counts, one for DVD and Live without uniq'ing the IP addresses doing the retrievals (because there could be multiple downloads from people behind firewalls), and one with. In both cases I think the numbers are very significantly higher than our current stats show. The raw numbers per day since F10 release are attached. -- 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

1 0

More auth options
by Dennis Gilmore 27 Apr '09

27 Apr '09

So doing a liitle looking around I cane across some options that look interesting, the following options would mean you need to physically have something to login. yubikey http://www.yubico.com/products/yubikey/ It would require a pam module and for us to setup a server for managing keys. it looks to be fairly low cost. it would implement a 2 facter authentication. etoken http://www.aladdin.com/etoken/devices/pro-usb.aspx it moves the public key from your hard drive to something you physically need to have ubikey is max USD$25 where the etoken is probably at least USD$30. I would think that with yubikey we could work out a deal with them to get a discount in return for us being a case study/prominent user of there product. all of the software for yubikey AFAICT is open source. some of it would require packaging. Dennis

7 7

deltarpms not working since rawhide was signed
by Chuck Anderson 27 Apr '09

27 Apr '09

As stated by Jonathan Dieter in the bug below, deltarpms are mucking up rawhide updates right now because the drpms were created before the packages were signed, and the signed versions don't match the deltarpm reconstructed versions. For me at least, this is causing a problem because I'm not using a mirrorlist right now (too many problems with metalink mismatches). So when yum fails to accept the drpm-patched package, the yum update just fails outright because there are no more mirrors to get the full updated package from. Is there anything that can be done on the infastructure side as proposed below? https://bugzilla.redhat.com/show_bug.cgi?id=497459 Comment #2 From Jonathan Dieter (jdieter(a)gmail.com) 2009-04-24 11:18:36 EDT (-) [reply] ------- This is not a deltarpm bug or a yum-presto bug, but rather an Infrastructure bug. The deltarpm was created before the target rpm was gpg signed. So it does indeed build to a valid rpm with exactly the same data as the downloaded rpm, but without the signature. Because it's not exactly the same file, yum refuses to use it and redownloads the full (signed) rpm (which is what it should do). The infrastructure should either delete and regenerate drpms after the rpm signatures have changed or they should use the code fragment from https://fedorahosted.org/koji/ticket/38#comment:3 to attach rpm signatures to deltarpms. Not sure how to reassign to Infrastructure.

8 15

App3 reboot (request)
by Mike McGrath 24 Apr '09

24 Apr '09

Can I get 2 +1's to reboot app3 (which is currently frozen) I'd like to power down xen13 and power it back up now that the BMC has been flashed. There shouldn't be any impact to the users except that /transifex/ will go down during the reboot. Which shouldn't be a problem as I believe most people are using /tx/ now anyway. 2 +1's ? -Mike

3 3

Meeting Log - 2009-04-23
by Ricky Zhou 23 Apr '09

23 Apr '09

19:59 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Meeting who's here? 19:59 * ricky (but will probably leave in the middle) 19:59 < ggruener> pong 20:00 * SmootherFrOgZ is 20:00 < mmcgrath> ok, lets get started 20:00 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Cloud Stuff 20:00 * nirik is the seats in the back. 20:00 -!- maploin [n=mapleoin@fedora/maploin] has joined #fedora-meeting 20:00 < mmcgrath> So the cloud stuff is coming along. Not much to report on an ETA for when we'll be giving out guests yet. 20:01 < mmcgrath> We've found (and filed) bugs that will probably need to be taken care of. 20:01 < mmcgrath> Generally though it's moving along 20:01 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Xen13 20:01 -!- dkovalsk [n=dkovalsk(a)ip-89-103-122-242.karneval.cz] has quit Client Quit 20:01 < mmcgrath> so xen13 still hasn't been fixed and IBM has been slow getting back to me about whether or not this is the type of thing our on site warranty will cover. 20:01 * johe around to (just to tell) 20:02 < mmcgrath> johe: hey 20:02 * abadger1999 here 20:02 * mmcgrath makes note to get back to them today, haven't heard back 20:02 < mmcgrath> We're change frozen for the release. 20:03 < abadger1999> was the fan on xen13 a red herring? 20:03 < mmcgrath> abadger1999: well it was fan 7 (or something) which the server didn't have 20:03 < mmcgrath> The techs wanted me to power it down, disconnect the cmos battery and wait 5 minutes 20:03 < mmcgrath> but I'm not there 20:03 < mmcgrath> and no one will be on site. 20:03 -!- knurd is now known as knurd_afk 20:03 < mmcgrath> they think something buggy is happening in the cmos. 20:03 < mmcgrath> quite odd 20:04 < mmcgrath> But the preview release is scheduled for the 28th currently 20:04 < mmcgrath> AFAIK it's on schedule 20:04 < mmcgrath> https://fedorahosted.org/fedora-infrastructure/report/9 20:04 < mmcgrath> looks like everything is assigned and ready 20:04 < mmcgrath> Anyone have any questions or comments on the open preview tickets? 20:06 < mmcgrath> k 20:06 < mmcgrath> So that's really all I had to discuss for thsi meeting 20:06 < mmcgrath> sorry its short but I've got a couple of things going on at once 20:06 < mmcgrath> just the same 20:06 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Open Floor 20:06 < mmcgrath> does anyone have anything to discuss? 20:07 < mmcgrath> If not I'll close the meeting in 30 :) 20:08 < johe> that was fast 20:08 < f13> I have.... nothing. 20:08 < mmcgrath> johe: yeah this was a quick one. 20:08 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Meeting Closed

1 0

FAS is way too slow.
by susmit shannigrahi 23 Apr '09

23 Apr '09

FYI. Right now, it is talking upto 3 minutes to log-in. Don't know why. Thanks. -- Regards, Susmit. ============================================= ssh 0x86DD170A http://www.fedoraproject.org/wiki/user:susmit =============================================

5 5

[MirrorManager PATCH] Fix killall path, add requires on psmisc for killall.
by Ricky Zhou 23 Apr '09

23 Apr '09

We recently got some cron mail about a wrong killall path, so here's a patch to mirrormanager to fix it. --- mirrormanager.spec.in | 2 +- server/logrotate.conf | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/mirrormanager.spec.in b/mirrormanager.spec.in index cc154b9..8c3d08f 100644 --- a/mirrormanager.spec.in +++ b/mirrormanager.spec.in @@ -10,7 +10,7 @@ URL: http://fedorahosted.org/mirrormanager Source0: https://fedorahosted.org/releases/m/i/%{name}/%{name}-%{version}.tar.bz2 BuildRoot: %(mktemp -ud %{_tmppath}/%{name}-%{version}-%{release}-XXXXXX) BuildRequires: python -Requires: TurboGears, python-IPy, python-GeoIP, wget, yum +Requires: TurboGears, python-IPy, python-GeoIP, wget, yum, psmisc %define py_ver %(echo `python -c "import sys; print sys.version[:3]"`) %if "%{py_ver}" < "2.5" diff --git a/server/logrotate.conf b/server/logrotate.conf index ec473e7..733de26 100644 --- a/server/logrotate.conf +++ b/server/logrotate.conf @@ -13,6 +13,6 @@ dateext rotate 15 postrotate - /sbin/killall -HUP crawler_perhost + /usr/bin/killall -HUP crawler_perhost endscript } -- 1.6.0.6 Thanks, Ricky

4 5

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

infrastructure April 2009