[12:09:10] <ggherdov`>	 Hello. Is there any (public) place where you people document how do you version database changes? There is being some discussion around db change versioning this days
[12:09:10] <ggherdov`>	 https://news.ycombinator.com/item?id=10034546 and https://www.reddit.com/r/Database/comments/3girji/database_versioning_best_practices/
[12:09:10] <ggherdov`>	 I work (tangentially) with databases and wanted to know some more about how complex organizations handle this. (Sorry if OT, I didn't intend to be annoying)
[12:10:55] <jynus>	 this could be a good place: https://www.percona.com/live/europe-amsterdam-2015/sessions/mysql-wikipedia-how-we-do-relational-data-wikimedia-foundation
[12:12:10] <jynus>	 at wikimedia we do nothing special, we version database schema along the rest of the code
[12:12:19] <jynus>	 check the git repo for mediawiki for more info
[12:16:31] <ggherdov`>	 Thanks jynus for the link, next month in europe. I could actually drop by.
[13:15:12] <jynus>	 ggherdov`, also https://www.percona.com/live/mysql-conference-2013/sites/default/files/slides/PLMCE13_online-migration.pdf
[13:15:34] <ggherdov`>	 oh nice
[13:16:02] <jynus>	 you may or may not like the specific tool, but lefred has very nice ideas
[15:23:01] <jynus>	 Krenair, production checksums look way better: https://phabricator.wikimedia.org/P1866
[16:37:33] <jynus>	 Ongoing things: 1) db1042 logically being reconstructed- being vslow/dumps it had an oversized ibdata1
[16:38:28] <jynus>	 2) running checksums in a subset of s7 tables. Works fairly well, but some tables cannot be checked due to being too large and not having proper unique keys (revision)
[18:00:43] <jynus>	 again, another tokudb table crashed on labs1003; rebuilt it
[19:09:23] <jynus>	 3) Fixing labs-enwiki slave drift
[19:10:54] <jynus>	 Because of the select...inserts on filearchive, some servers on s4 are a bit degraded rows-wise; will checksum s4 next
[19:11:33] <jynus>	 once checksums finish on s7, will fix dbstore2x