Build:OutageReports:20070410-02: Difference between revisions

From MozillaWiki
Jump to navigation Jump to search
No edit summary
 
 
Line 1: Line 1:
== Outage Template ==
== Outage Template ==


On Apr 10th from ~ 4 am, moz180-linux-tbox experience a service outage for ~ 10 hours.
On Apr 10th from ~ 4 am, moz180-linux-tbox experienced a service outage for ~ 10 hours.


'''What was affected:'''
'''What was affected:'''
Line 14: Line 14:


Some ideas
Some ideas
* Monitor tinderbox disk space (eg Nagios)
* Monitor tinderbox disk space (eg Nagios)
* Make tinderbox go red on symbol problems (creation and push)
* Make tinderbox go red on symbol problems (creation and push)

Latest revision as of 21:26, 10 April 2007

Outage Template

On Apr 10th from ~ 4 am, moz180-linux-tbox experienced a service outage for ~ 10 hours.

What was affected:

Firefox linux builds on Mozilla1.8.0 branch (en-US and locales)

What was the cause of the outage:

The /builds partition was full. It looks like symbol generation has been busted for the last few days (at least), with firefox-bin.elf ending up a random, truncated, size and zero byte files for the other executables.

What will be done to prevent this in the future:

Some ideas

  • Monitor tinderbox disk space (eg Nagios)
  • Make tinderbox go red on symbol problems (creation and push)