1,107
edits
Webdev:Meetings:2010-09-28: Difference between revisions
Jump to navigation
Jump to search
Webdev:Meetings:2010-09-28 (view source)
Revision as of 17:08, 28 September 2010
, 28 September 2010→laura
m (→wil) |
(→laura) |
||
| Line 27: | Line 27: | ||
== lars == | == lars == | ||
== laura == | == laura == | ||
* Firefighting: | |||
** After last webdev meeting, rest of week was consumed with 1.8 [https://intranet.mozilla.org/Socorro:1.8_Postmortem release/disaster/rollback/postmortem] | |||
** Last week began with a day off and then [https://intranet.mozilla.org/Socorro:1.8_Postmortem#Outcomes_.2F_Actions grand plans to make things better] | |||
** and ended with further disasters (consuming Thursday and Friday) due to a hole in an HBase table, which we coded around with duct tape and then later fixed thanks to Cloudera Support and help from our friends at StumbleUpon. | |||
** This week began with a new disaster, which an entire day of work revealed to be due to a failed hard drive on cm-hadoop06. Yes, Hadoop is meant to be redundant if it loses a node. As long as it's not the wrong node. And no, monitoring did not alert us to the drive going bad. | |||
== lorchard == | == lorchard == | ||
== malexis == | == malexis == | ||