Final Update: Thursday, October 25th 2018 09:24 UTC
We’ve confirmed that all systems are back to normal as of 7:50 UTC. Our logs show the incident started on 6:15 UTC and that during the 1 hour and 35 minutes that it took to resolve the issue, customers might have observed slowness in Central US.Sorry for any inconvenience this may have caused.
- Root Cause: The initial symptoms indicate dip in available memory for one of the web tier node.
- Chance of Re-occurrence: High - We are still investigating what caused the High memory utilization.
- Lessons Learned: While we don't fully understand what caused the dip in available memory, but we do understand that recycling the web tier node mitigates the issue.
- Incident Timeline: 1 hour & 45 minutes – 06:15 UTC through 07:50
Initial notification: Thursday, October 25th 2018 07:55 UTC
We're investigating a possible customer impacting event in Central US. We are triaging the issue and will provide more information as
it becomes available.
- Next Update: Before Thursday, October 25th 2018 08:30 UTC