DB performance of frequently updated table
by Michael Šimáček
Hi,
We've been facing some DB performance issues in Koschei production
machine recently. Our central table (package) has only ~10000 rows but
sequential scan of the table was taking unreasonably long (~4s). Other
tables that are orders of magnitude bigger were faster to query. I was
investigating the problem and it turned out that the table occupied very
large amount of disk space:
koschei=> analyze verbose package;
INFO: analyzing "public.package"
INFO: "package": scanned 30000 of 322434 pages, containing 1086 live
rows and 12973 dead rows; 1086 rows in sample, 35640 estimated total rows
Regular vaccum didnt help much, but vacuum full on the table returned
the performance back to normal (and slowly degrades again, so currently
we just periodically vaccum full).
My question is how could it get into such poor condition (over like 2
days) when almost no code that manipulates the table was changed since
the version that was deployed in cloud and was running fine for months?
The table is frequently updated (recalculating the package priority
every few seconds), is there some difference in configuration of
[auto]vacuuming in the production DB from postgres defaults (which were
used on cloud instance)?
Thanks,
Michael Simacek
8 years, 9 months
[release] pagure: 0.1.19
by Pierre-Yves Chibon
Good morning everyone,
After almost a month of development, I just released a new version of pagure:
0.1.19
The list of changes in this version if pretty big (details following) but I
think 2 of the biggest changes are:
a) the doc server is now a separate wsgi application, thus solving any potential
cross-site forgery we could have ran into by displaying content submitted by our
users
b) projects can be tagged and search in the API using them, meaning we can have
an `fedora-infra` tag for projects relevant to the Fedora Infra and use this tag
for notifications (FMN, IRC & co).
But here is the full changelog:
* Mon Jul 20 2015 Pierre-Yves Chibon <pingou(a)pingoured.fr> - 0.1.19-1
- Update to 0.1.19
- Prettify the JSON stored in the git for tickets/requests... (Simo Sorce)
- Use the project name as subject tag in the notifications sent (Simo Sorce)
- Add an X-pagure header with either the pagure instance or the project name
- Reset the merge status of all the open PR when one is merged
- Add a second server listing the number of connections opened on the first
eventsource server
- Log the info instead of printing them in the eventsource server
- Split the documentation to a different wsgi application to avoid any risk of
cross-site forgery
- Fix the JS logic when adding a tag or a dependency to avoid having duplicates
in the input field
- Allow deleting a git branch of a project via the UI
- Include the font-awesome in the source rather than relying on an external cdn
- Do not try to connect to the eventsource server if we're not viewing a
pull-request
- Fix showing the first comment made on a PR via the eventsource server
- Fix showing the git URLs in the doc server
- Much better API documentation (Lei Yang)
- Handle showing closed PR that were not merged
- Fix refreshing the UI of private tickets via the eventsource (making calls to
the API to get the info while only getting what changed via the SSE)
- Fix the anchor links in the API documentation
- Blink the tab upon changes in the page
- Ensure we close both SSE server when stopping pagure_ev
- Let the HTML form trigger if we did not connect to the EV server successfully
- The admins of a repo are anyone with commit access to the repo, directly or
via a group
- Order the project by names in the front page (instead of creation date)
- Add the ability to tag a project
- Fix the fedmsg_hook when there are only deletions or only additions
- Add a new API endpoint allowing to search projects (by name, author, tag ...)
- Make pagure compatible with pygit 0.22.0
- Adjust unit-tests for all these changes
It is currently running in stg, I will let it here a day or two before updating
prod and with some luck maybe we will have the ssl cert ready by then.
Happy testing!
Pierre
8 years, 9 months
Plan for tomorrow's Fedora Infrastructure meeting - 2015-07-23
by Kevin Fenzi
The infrastructure team will be having it's weekly meeting tomorrow,
2015-07-23 at 18:00 UTC in #fedora-meeting on the freenode network.
We have a gobby document
(see: https://fedoraproject.org/wiki/Gobby )
fedora-infrastructure-meeting-next is the document.
Please try and review and edit that document before the meeting and we
will use it to have our agenda of things to discuss. A copy as of today
is included in this email.
If you have something to discuss, add the topic to the discussion area
with your name. If you would like to teach other folks about some
application or setup in our infrastructure, please add that topic and
your name to the learn about section.
kevin
--
= Introduction =
This shared document is for the next fedora infrastructure meeting.
We will use it over the week before the meeting to gather status and info and
discussion items and so forth, then use it in the irc meeting to transfer
information to the meetbot logs.
= Meeting start stuff =
#startmeeting Infrastructure (2015-07-23)
#meetingname infrastructure
#topic aloha
#chair smooge relrod nirik abadger1999 lmacken dgilmore mdomsch threebean pingou puiterwijk pbrobinson
#topic New folks introductions / Apprentice feedback
#topic GSoC student update - kushal
= Status / information / Trivia / Announcements =
(We put things here we want others on the team to know, but don't need to discuss)
(Please use #info <the thing> - your name)
#topic announcements and information
#info cloud migration moving along, possibly done end of next week - kevin/patrick
#info mass updates/reboots done. kevin/patrick/relrod/smooge/tflink
#info openstack update again a nightmare, but patrick got it working - patrick
#info production bodhi2 instances ready to go - kevin
#info
= Things we should discuss =
We use this section to bring up discussion topics. Things we want to talk about
as a group and come up with some consensus or decision or just brainstorm a
problem or issue. If there are none of these we skip this section.
(Use #topic your discussion topic - your username)
# Centralized logging - jcvicelli
https://fedorahosted.org/fedora-infrastructure/ticket/3627
looks like it was all done, should we close this ticket?
#topic
= Learn about some application or setup in infrastructure =
(This section, each week we get 1 person to talk about an application or setup
that we have. Just going over what it is, how to contribute, ideas for improvement,
etc. Whoever would like to do this, just add the info in this section. In the
event we don't find someone to teach about something, we skip this section
and just move on to open floor.)
#topic Learn about:
= Meeting end stuff =
#topic Open Floor
#endmeeting
8 years, 9 months
[release] MirrorManager2: 0.4.0
by Pierre-Yves Chibon
Good morning everyone,
I just released a new version of MirrorManager2: 0.4.0
Here is its changelog:
* Wed Jul 22 2015 Pierre-Yves Chibon <pingou(a)pingoured.fr> - 0.4.0-1
- Update to 0.4.0
- Add utility script to update the MM2 DB for a single file (light-weight UMDL
for a single file) (Ralph Bean)
- Add support to gracefully shutdown the crawler (Adrian Reber)
- Add support to limit crawling by continent (Adrian Reber)
- Fix the crawler to work properly with MM2
- Fix the repomap to create properly development repos
- Fix new repository detection and creation in UMDL (Adrian Reber)
All the changes are related to the backend part, either the crawler or the UMDL
(and Ralph's utility script).
So I'll let it stay in stg for a little bit and will likely push the update to
prod later today or tomorrow.
Pierre
8 years, 9 months
Installing fedora-packager on arm-packager hosts
by Mikolaj Izdebski
I think that arm-packager machines are useful to debug build failures
and other packaging issues specific to ARM or armhfp.
One problem is that default installation of these hosts is missing basic
packager tooling (fedora-packager with dependencies). Even if I install
these tools myself, changes are lost after given instance is recreated.
Moreover, current policy requires reverting all changes made when they
are no longer needed.
Therefore I propose to:
1) install fedora-packager package on arm-packager hosts
2) adjust mock PAM config (/etc/pam.d/mock) to allow members of packager
group to use mock without having to add themselves to mock group
Any comments or objections?
--
Mikolaj Izdebski
Software Engineer, Red Hat
IRC: mizdebsk
8 years, 9 months
Meeting Agenda Item: Introduction of Joshua Hoblitt
by Joshua hoblitt
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Hi Folks,
I'm interested in getting involved with Fedora Infrastructure.
Primarily with
the build tooling (koji, copr, etc) -- there is some congruence with my
current $day_job tasking of migrating a fairly complex stack to jenkins
for CI and developing a packaging pipeline. I'm also generally
interested in metrics collection and monitoring.
A longer term area I'm interested in exploring is convergence between CI
configuration (.travis.yml, jenkins job configuration, etc.) and
packaging specifications. More immediately, I'm intereseted in working
on CM/automation of any of the distributed systems (including but
certainly not limited to koji).
A bit about my background: I'm fairly strong in configuration management
(largely the puppet ecosystem and theforeman), both user space (home grown,
iRODS) & distributed (GPFS) filesystems, and modest scale HPC. Like
most people, I have my personal technology preferences but try to
pragmatic about tooling. I've been working in ruby, python, node, java,
and groovy [thus far] this year. I'm also currently a vagrant/terraform
addict.
I'm looking forward to working with such an awesome community [if it
will have me ;)]!
Cheers,
- -Josh
Ps. Fedora's usage of yubikeys has inspired me to get one for my
personal use -- its arriving tomorrow.
- --
irc: ^conner
https://github.com/jhoblitt
http://linkd.in/jhoblitt
https://forge.puppetlabs.com/jhoblitt
https://rubygems.org/profiles/jhoblitt
http://search.cpan.org/~jhoblitt/
https://joshua.hoblitt.com/rtfm/
https://joshua.hoblitt.com/resume/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQIbBAEBCAAGBQJVqdfeAAoJEOoY5YqfEKu+CqUP93exJtl5SrBupG67AKzWxCzI
tnNqaUKurYzS+c2/sDLOzs6giTZzz08rQ9e5okmofpH3RP1Q6pzf567Y8F0evGgB
ms+7gZ9tci4qiw6FfVNUCBp1ETS16Wnqt7EEioi2tx7j+jRaRe2/C8hdiZnhEiDD
fYx9H8D/C2amSVQY8l/UMNiaHVzx/AP1LBwVrR986t1NptqYuTVPY9bTzrJvze3W
NZV5uVlBFGPrb7bOj1dOteLU3OagsUp1fbpYqdohzF1EsX6deQXKOy0AjMvP/R82
1QNiLe0KLqEnMKVS3SkjRUJOucDK3k40YG7bOK8RWfSRT9wWMRWVBvCLMxEPE3y3
rCbvwpbWoODVfdlzizTKO2+Or64Az7ZpdVfhAUbxZleMY06N4cGq4OtK26ERrUwM
tiWUtutqdHIuRVYCCBh+x0Ltt5TEBtksCYjNwoNR+E6Ut5A7iy033qNBPQ5Jy3kE
zH4Rig18M4QYk1xqyg9QFCJrZ5x1CXAbU4wAUgxFE/1qw/03MtaPRIoAs1EniYSO
FxJQyySLPOOow54g/vjsxqJyhvzFd4Xk6pQ9mgx9YCrbjNXbLtts7mj6jEtC7csS
uwdvI2RRPPCJxqXN0bMkiqAiJHx7kEkdKRcuzIvgN6KTRe5CvgzW/XPtcQQazkI8
R/6F5tGK+6uXYqY0G1c=
=squF
-----END PGP SIGNATURE-----
8 years, 9 months
Plan for tomorrow's Fedora Infrastructure meeting - 2015-07-16
by Kevin Fenzi
The infrastructure team will be having it's weekly meeting tomorrow,
2015-07-16 at 18:00 UTC in #fedora-meeting on the freenode network.
We have a gobby document
(see: https://fedoraproject.org/wiki/Gobby )
fedora-infrastructure-meeting-next is the document.
Please try and review and edit that document before the meeting and we
will use it to have our agenda of things to discuss. A copy as of today
is included in this email.
If you have something to discuss, add the topic to the discussion area
with your name. If you would like to teach other folks about some
application or setup in our infrastructure, please add that topic and
your name to the learn about section.
kevin
--
= Introduction =
This shared document is for the next fedora infrastructure meeting.
We will use it over the week before the meeting to gather status and info and
discussion items and so forth, then use it in the irc meeting to transfer
information to the meetbot logs.
= Meeting start stuff =
#startmeeting Infrastructure (2015-07-16)
#meetingname infrastructure
#topic aloha
#chair smooge relrod nirik abadger1999 lmacken dgilmore mdomsch threebean pingou puiterwijk pbrobinson
#topic New folks introductions / Apprentice feedback
#topic GSoC student update - kushal
= Status / information / Trivia / Announcements =
(We put things here we want others on the team to know, but don't need to discuss)
(Please use #info <the thing> - your name)
#topic announcements and information
#info Various fixes to people01 after migration - kevin
#info Outage template updated - kevin
#info lots of cloud instances moved to new cloud - kevin
#info lots of no longer needed cloud instances killed in old cloud - kevin
#info osbs01.stg setup for releng - kevin
#info backups migrated for hosted04->03 and ongoing for collab04->03 - kevin
#info inode count increased for backups volume when it hit 100% - kevin
= Things we should discuss =
We use this section to bring up discussion topics. Things we want to talk about
as a group and come up with some consensus or decision or just brainstorm a
problem or issue. If there are none of these we skip this section.
(Use #topic your discussion topic - your username)
#topic
= Learn about some application or setup in infrastructure =
(This section, each week we get 1 person to talk about an application or setup
that we have. Just going over what it is, how to contribute, ideas for improvement,
etc. Whoever would like to do this, just add the info in this section. In the
event we don't find someone to teach about something, we skip this section
and just move on to open floor.)
#topic Learn about: backups with rdiff-backup - kevin
= Meeting end stuff =
#topic Open Floor
#endmeeting
8 years, 9 months
Meeting Agenda Item: Introduction Joshua Santos
by Joshua Santos
I had previously expressed interest in working with the fi-apprentice group
but had a lot of personal issues come up which have been all sorted out and
I can now devote time to assisting with this :) I will be attending the
meeting this Thursday at 6:00 PM UTC
IRC Handle: nerdsville
Skills: Django, Python, Ruby, C#, PHP, Apache, Networking
I had worked previously with Pingou and he had a task for me on the
election side of things, which I will be working on between now and Thursday
On the infra side of things, looking at the outstanding issues, I am
interested in
#4744 <https://fedorahosted.org/fedora-infrastructure/ticket/4744> implement
account recovery
<https://fedorahosted.org/fedora-infrastructure/ticket/4744>
#2888 <https://fedorahosted.org/fedora-infrastructure/ticket/2888> Implement
HTTP Strict Transport Security
<https://fedorahosted.org/fedora-infrastructure/ticket/2888>
Talk to you all on Thursday! :)
--
Regards,
Joshua Santos
CTO/Founder
Nerdsville LLC
8 years, 9 months
Introducing myself to the Fedora infrastructure team
by satej kumar Sahu
Hi,
I am new to fedora project and have joined the infrastructure team. I
am a web developer by profession and would like to contribute to this
team.
Please guide me. Looking forward to collaborating with you all.
Thanks,
Satej
8 years, 9 months
