*bump* (there was already a reminder about this in Scrum of
Scrums recently, but another one here can't hurt)
To make it extra clear, it's not enough to whitelist information on the
schema talk page, filing a Phabricator ticket or such is needed too. The
most recent version of the whitelist (here
<https://gerrit.wikimedia.org/r/#/c/298721/5/files/mariadb/eventlogging_purging_whitelist.tsv>,
temporary location) dates from July 2016, and while some tickets may have
been filed recently, it's quite possible that there are people who are not
aware of this requirement. Here is a Quarry query for all schema pages (and
talk pages) that have been edited since 2016:
https://quarry.wmflabs.org/query/19014 - it might be worth taking a look to
see whether a schema you use appears there and if yes, if these edits
necessitate whitelist updates.
BTW, for the record, the first step of this purging effort already
happened, involving <https://phabricator.wikimedia.org/T161855> dropping
completely 124 old EventLogging tables that had no event recorded in the
last 90 days and were not in the whitelist. A list
<https://gist.github.com/ottomata/df9d4615b8ffdf538faf6a005683e1fc> of
these tables that were deleted last month is now available.
On Mon, Apr 3, 2017 at 1:10 PM, Nuria Ruiz <nuria(a)wikimedia.org> wrote:
Team:
This is just a friendly remainder regarding the fact that if you want to
preserve data (of non private nature) in eventlogging beyond the 90 period
you have to let us know via ticket or similar. The info per schema should
be available on each schema's talk page.
Purging info can be found here:
https://wikitech.wikimedia.org/wiki/Analytics/EventLogging/
Data_retention_and_auto-purging
Thanks,
Nuria
_______________________________________________
Analytics mailing list
Analytics(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/analytics
--
Tilman Bayer
Senior Analyst
Wikimedia Foundation
IRC (Freenode): HaeB