Talk:Incident documentation/20160915-MediaWiki

From Wikitech
Jump to navigation Jump to search

Hashar: Thank you for working on this issue. I would go even further concerning alarms for the account creation error rate. It seems to me that having a 1% (not 100%) account creation error rate for all wikis, or a 2% error rate for any single wiki, should be cause for investigation. It might emerge, for example, that 100% of the users who are trying to create accounts with a particular browser are unable to do so. I'm not sure if this would qualify as a "critical alarm" but I feel that it would warrant a rapid investigation, even at a low threshold. --Pine (talk) 19:13, 16 September 2016 (UTC)

Add monitoring alarm for Account creation errors is part of phab:T140942. The shell out fails to capture standard error. also deserves its own action item IMO; I'm sure the output could be collected to a tempfile and then logged. Chad reported the error on IRC on Tuesday and a few of us looked at it (without realizing the severity) and probably gave up on investigating after seeing that the error message just disappears in the void (I did, at least). --tgr (talk) 23:24, 16 September 2016 (UTC)