Ebuild request for PgBouncer, which is a Lightweight connection pooler for PostgreSQL. == Features == * Several levels of brutality when rotating connections: Session pooling:: Most polite method. When client connects, a server connection will be assigned to it for the whole duration it stays connected. When client disconnects, the server connection will be put back into pool. Transaction pooling:: Server connection is assigned to client only during a transaction. When !PgBouncer notices that transaction is over, the server will be put back into pool. This is a hack as it breaks application expectations of backend connection. You can use it only when application cooperates with such usage by not using features that can break. See the table below for breaking features. Statement pooling:: Most aggressive method. This is transaction pooling with a twist - multi-statement transactions are disallowed. This is meant to enforce "autocommit" mode on client, mostly targeted for PL/Proxy. * Low memory requirements (2k per connection by default). This is due to the fact that !PgBouncer does not need to see full packet at once. * It is not tied to one backend server, the destination databases can reside on different hosts. * Supports online reconfiguration for most of the settings. * Supports online restart/upgrade without dropping client connections. * Supports protocol V3 only, so backend version must be >= 7.4. Reproducible: Always
Created attachment 169434 [details] ebuild for version 1.2.3 works on PSQL-8.4.3 on x86 and sparc.
Created attachment 169436 [details, diff] auto-database patch patch to avoid having to manually enter database servers.
Created attachment 169456 [details] ebuild for version 1.2.3 few changes added debug USE
Created attachment 169484 [details] updated for new init.d script
Created attachment 169486 [details] init.d script
Hey! Thanks a bunch for your ebuild. I would really like to see this in the tree as well, not sure about the patch though. I think similar feature enhancements should go through the "normal procedure" - sending a patch to the pgbouncer developers and let it merge upstream (which is on its way as far as the patch info on pgfoundry tells me) A couple of minor remarks from my point of view: (* Remove the auto-database.patch) * How about wrapping --enable-cassert with the debug flag? * Your indentation in both init.d and the ebuild is a bit messed up * add the ~amd64 keyword (works fine here with postgresql 8.3.4) * Don't remove the pid in start() - that's what Gentoo's init system is for * Do we really want to use the postgres user for pgbouncer? postgres owns a lot of files that pgbouncer has nothing to do with - as a proxy, I'm inclined to let it sit in its own user/group Again, this is a very nice addition to portage - thanks for your time!
(In reply to comment #6) > A couple of minor remarks from my point of view: > (* Remove the auto-database.patch) > * How about wrapping --enable-cassert with the debug flag? > * Your indentation in both init.d and the ebuild is a bit messed up > * add the ~amd64 keyword (works fine here with postgresql 8.3.4) > * Don't remove the pid in start() - that's what Gentoo's init system is for > * Do we really want to use the postgres user for pgbouncer? postgres owns a > lot of files that pgbouncer has nothing to do with - as a proxy, I'm inclined > to let it sit in its own user/group > > Again, this is a very nice addition to portage - thanks for your time! > > The Upstream knows about the patch, they have taken a look at it: http://pgfoundry.org/tracker/index.php?func=detail&aid=1010460&group_id=1000258&atid=985 Yea i agree, about wrapping the debug flags into 1 use, i'll do that. Yea i noticed that indentation issue on my other computer, i'll have to re-do it there. I'll take a look at the init script, but i think on one of my servers i had an issue with pid files. The reason i ran it as postgres user, is that it saves a lot of tinkering with permissions( a few a+rw) , configuration, and it makes it not very "working out of the box" esk, but that just my opinion. Since there isn't a lot of documentation of this software, i thought keeping it simple would make it more likely for people to want to use it. I'll submit an updated ebuild, when i am finished revising it. -Mike
Created attachment 174004 [details] updated ebuild updated the ebuild and init.d script based on suggestions.
Created attachment 174005 [details] init.d script
Created attachment 174006 [details] Updated Ebuild
Created attachment 187023 [details] pgbouncer 1.3 Pgbouncer 1.3 has been out a while. Updated ebuild with following changes: * version bump * changed download path (can't figure out a generic pgfoundry download path) * removed commented out patch (merged upstream) * changed depend to virtual/postgresql-base since -server on the same computer isn't needed. * changed name of init script to ${PN} (skipping version) * remove eautoreconf since i didn't have any autotools issues on any of my testbeds - please re-add if necessary The init file should be renamed pgbouncer-1.3 ; otherwise its good to go. I guess one could move some minor things from the init script to conf.d or something - but it works as intended.
(this is an automated message based on filtering criteria that matched this bug) 'EBUILD' is in the KEYWORDS which should mean that there is a ebuild attached to this bug. This bug is assigned to maintainer-wanted which means that it is not in the main tree. Heuristics show that no Gentoo developer has commented on your ebuild. Hello, The Gentoo Team would like to firstly thank you for your ebuild submission. We also apologize for not being able to accommodate you in a timely manner. There are simply too many new packages. Allow me to use this opportunity to introduce you to Gentoo Sunrise. The sunrise overlay[1] is a overlay for Gentoo which we allow trusted users to commit to and all users can have ebuilds reviewed by Gentoo devs for entry into the overlay. So, the sunrise team is suggesting that you look into this and submit your ebuild to the overlay where even *you* can commit to. =) Because this is a mass message, we are also asking you to be patient with us. We anticipate a large number of requests in a short time. Thanks, On behalf of the Gentoo Sunrise Team, Jeremy. [1]: http://www.gentoo.org/proj/en/sunrise/ [2]: http://overlays.gentoo.org/proj/sunrise/wiki/SunriseFaq
Created attachment 190924 [details] Updated ebuild New version which fixes all warnings from repoman
This ebuild is now in the gentoo-sunrise overlay: http://overlays.gentoo.org/proj/sunrise/browser/sunrise/dev-db/pgbouncer
I would like to move this ebuild to the main tree. For me this has been a lifesaver multiple times, and out of the box usually gives a nice performance improvement (see for instance http://filip.rembialkowski.net/pgbouncer-mini-howto-benchmark/ ). Would the postgres herd possibly be interested in picking it up (i will gladly act as co-maintainer)? Thanks
(In reply to comment #15) > I would like to move this ebuild to the main tree. For me this has been a > lifesaver multiple times, and out of the box usually gives a nice performance > improvement (see for instance > http://filip.rembialkowski.net/pgbouncer-mini-howto-benchmark/ ). > > Would the postgres herd possibly be interested in picking it up (i will gladly > act as co-maintainer)? > > Thanks Yes, we would. Contact me via IRC or email and we'll get the ball rolling.
Created attachment 292515 [details] pgbouncer-1.4.2.ebuild Fixes: - EAPI=4 - Clean up deps - Switch to the postgres src_uri mirrors None of this would of course happen if Aaron didn't help out. Thanks!
Created attachment 292517 [details] pgbouncer.initd Changes: - Create rundir if it doesn't exist - Switch from opts to extra_started_commands Again: Thanks to Aaron W. Swenson!
16 Nov 2011; Aaron W. Swenson <titanofold@gentoo.org> +pgbouncer-1.4.2.ebuild, +files/pgbouncer.initd, +metadata.xml: Initial commit. Ebuild written by Johan Bergström <bugs@bergstroem.nu>. Fixes bug 243232.
