GRAVENHURST Incident

Incident Report for WPCloud

Postmortem

Sites on the GRAVENHURST cluster had several brief interruptions between Tuesday 3 November and Saturday 7 November. The cause is a hardware fault in the physical server that hosts the cluster. We are moving the cluster to separate replacement hardware so the original server can be repaired or replaced.

What happened

The physical host server stopped responding four times:

  • Tuesday 3 November, about 8:18 a.m. ET (a few minutes)
  • Wednesday 4 November, about 11:25 p.m. ET (up to about 20 minutes)
  • Saturday 7 November, about 7:25 a.m. ET (a few minutes)
  • Saturday 7 November, about 7:35 a.m. ET (a few minutes)
    Each time, the cluster restarted and sites came back online. Site databases recovered automatically after each restart, and our integrity checks this morning found no damage.

Cause

The host server recorded processor-level hardware errors. After the first event, we applied the standard software fix for this type of error. The fault returned with the fix in place, which confirms a hardware problem with the server itself, not with the cluster's software or configuration.

What we are doing

  • A complete copy of the cluster is now on separate replacement hardware and is kept up to date every few minutes.
  • We will switch the cluster to the replacement hardware on Sunday 8 November at 6:00 a.m. ET. Sites may be unavailable for a few minutes during the switch.
  • If the original server fails again before then, we will switch immediately instead.
  • The original server will then go to our data centre provider for inspection and repair or replacement.
  • Once the original hardware is confirmed healthy, we may move the cluster back during a planned maintenance window, with advance notice.

No action is needed from you.

Posted Nov 07, 2026 - 14:35 EST

Resolved

Server remains stable at this time. Sysadmin team will continue to monitor & will post more details in port mortem if available.
Posted Nov 07, 2026 - 07:46 EST

Update

Server is now responding our sysadmin team continues to investigate.
Posted Nov 07, 2026 - 07:38 EST

Investigating

We are currently investigating an issue impacting the GRAVENHURST cluster.
Posted Nov 07, 2026 - 07:33 EST
This incident affected: MONTREAL DATA CENTRE.