Service problems, 11 Dec

You may have noticed that we have had a downtime of about 45 minutes. All systems should now be back up and functioning normally.

The cause was scheduled maintenance which did not go as planned, and is related to the earlier outage this previous weekend.

In summary:

On Sunday:

  • Scheduled network maintenance caused an unexpected panic and reboot of one of our two redundant storage nodes.
  • Storage automatically failed over to the other storage node, but a misconfiguration meant that that our servers were not able to access the storage on the other storage node
  • Due to another misconfiguration, our monitoring/alerting systems did not detect the problem, and so remained undetected for several hours
  • After investigation, we were advised to replace the motherboard of the failed storage node by our vendor


  • The replacement of the motherboard took place, which should not have caused any downtime due to the redundant node
  • For an undetermined reason, the failover yet again failed to work (in a different way to before) and mbed had to stay down while the motherboard was replaced.

We will be investigating the root causes of these problems with our in house storage specialists and the vendor to avoid a recurrence of the problem.

Apologies for the inconvenience, and I will update this post with more as I know it.

1 comment on Service problems, 11 Dec:

13 Dec 2013

Thank you for keeping us updated.

You need to log in to post a discussion