Reactive Maintenance (Scheduled)
  • Priority - Medium
  • Affecting Server -
  • Summary

    Following this morning's issue with tempo, we have completed our investigation and will be performing maintenance on that single server.


    This evening, March 20th 2016, starting at 5pm Pacific Time.

    How Phast?

    The maintenance window will last a maximum of two hours but should take about twenty minutes.


    We've identified a condition that causes an IOPS bottleneck on tempo's main drive and are upgrading it from magnetic storage to an SSD for better throughput and performance. During this time we will take tempo offline to create snapshot of the current drive within Amazon Web Services, create a new SSD from it, swap the virtual drives, and boot tempo back up.

    Ongoing Benefit

    Tenants of our tempo server will notice an increase in performance each day following this upgrade.


    This is a no-to-low risk routine we perform on a regular basis. No data will be lost as a result of this work.

  • Date - 2016/03/20 17:00 - 2016/03/20 19:00
  • Last Updated - 2016/03/20 23:20