Before someone emails me about this... :-)
s1-analytics-slave eventlogging replication is starting to lag again (enwiki replication is ok).
I noticed that new eventlogging tables are using InnoDB instead of TokuDB on that slave. The issue is being fixed and we should be back up to speed within the day.
-s
Thanks for the heads up, Sean.
The VE graphs for enwiki [1] are currently down, all the other ones are up (see for example [2]) so it looks like there’s a problem with s1 that is not related to EventLogging (the VE scripts pull data from the MW database).
Dario
[1] http://ee-dashboard.wmflabs.org/dashboards/enwiki-metrics#edits-graphs-tab [2] http://ee-dashboard.wmflabs.org/dashboards/frwiki-metrics#edits-graphs-tab
On May 28, 2014, at 10:25 AM, Sean Pringle springle@wikimedia.org wrote:
Before someone emails me about this... :-)
s1-analytics-slave eventlogging replication is starting to lag again (enwiki replication is ok).
I noticed that new eventlogging tables are using InnoDB instead of TokuDB on that slave. The issue is being fixed and we should be back up to speed within the day.
-s _______________________________________________ Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics