On February 4, 2019, at approximately 11:35 PM PST, the xMatters monitoring tools alerted Client Assistance to an issue affecting service in the North American region. During the incident, some clients may have experienced some brief interruptions or delays, including 503 errors, when attempting to access the xMatters web user interface. While the underlying issue required approximately 12 hours to completely resolve, there was a total of 12 minutes of actual impact to clients. These impacts were spread across the incident duration in short intervals while the underlying issues were resolved.
The issue was related to unexpected and increased connection pool usage within the xMatters platform, which caused the web user interfaces and some API services to reach capacity and auto-heal multiple times. Increased query times on some databases resulted in back pressure on user-facing services; this decrease of performance resulted in connection pools reaching capacity.
As soon as the issue was detected, the Client Assistance team immediately initiated the internal Major Incident Management process and launched an investigation. The incident response team quickly identified an issue impacting the web user interface and declared a Severity 1 incident while engaging additional subject matter experts. Their first priority was to mitigate any client impact, and then work to identify a root cause and build a solution. When the issue reoccurred on February 5 at approximately 7:24 AM PST, the teams were able to immediately isolate the affected components and isolate the problematic services to perform remediation. They confirmed that this had correctly mitigated the problem, and that all services had been restored.
To prevent this issue from reoccurring, we are adding additional monitoring that will allow us to detect these types of incidents much earlier and automatically implement additional self-healing processes for affected service. In addition, we have conducted a thorough post-mortem and identified multiple areas where system resiliency can be improved.
January 04, 2019 - 23:35 - Monitoring tools alert Client Assistance to an issue in the North American region. Brief interruptions are detected.
January 04, 2019 - 23:40 - xMatters Client Assistance initiates major incident management process, launches investigation.
January 05, 2019 - 00:08 - Interruptions are no longer occurring.
January 05, 2019 - 07:24 - Second service impact begins (brief interruptions continue until 11:31)
January 05, 2019 - 08:04 - Status page updated to investigating: https://status.xmatters.com/incidents/t86p7lvvdn5g
January 05, 2019 - 08:05 - Status page updated to identified, incident team works to determine root cause and resolution options.
January 05, 2019 - 11:31 - Resolution initiated, status page set to monitoring.
January 05, 2019 - 12:42 - Production environment is determined to be stable, no further impact detected. Status page updated to resolved.
If you have any questions, please visit http://support.xmatters.com