At 6:17pm this evening, one of our core switches unexpectedly restarted.
Upon investigation, this was caused by a software fault in Cisco IOS which triggered an automatic restart and crashdump.
We will analyse this to see if we anticipate any further restarts of this switch.
Although no customers are directly connected to this core switch, it did cause a reconvergence of our network topology which may have caused a period of instability to customers.
No further instability is expected, but we will continue to monitor the device this evening.
All self-tests and internal monitoring of the device report no further issues.
The switch caused a larger reconvergence than expected as it was the master switch for our ring. This triggered a full reconvergence instead of a minor one.
The switch firmware crash occurred at 6:17pm, and rebooted and came back up at 6:21pm.
Some further instability would continue for a few minutes whilst BGP routing updated.