Beginning on Tuesday, October 9, 2018 at approximately 7:40 PM PST, the xMatters monitoring systems alerted Client Assistance to a potential issue with xMatters On-Demand services for clients in the North America region. During the incident, some customers may have experienced delays of up to 15 minutes in notification delivery. No notifications were lost during this period, and event injection and responses continued processing as normal.
This issue was caused by a previously unidentified defect within a service responsible for handling notification processing, triggered by an unusually high volume of notification requests.
As soon as the automated monitoring tools alerted xMatters Client Assistance to a possible delay in notification delivery, the teams began attempting to both reproduce the problem and determine the cause of the issue. Once the issue was confirmed, xMatters Client Assistance escalated the issue to Severity 1 and initiated the internal major incident management process. The incident response teams began working to identify and isolate the issue and quickly identified a problem with the notification service. The team discovered that some back-end services were in the process of automatically recovering from a failure and restarted one of the affected components to speed the recovery process. This appeared to resolve the issue and all services were restored. The teams concluded the major incident process, while continuing to monitor the situation and were able to identify the root cause as a defect in the notification service.
To prevent this issue from recurring, the Engineering team will upgrade to a newer version of the affected back-end service which contains a fix for the defect that caused the delay in processing notifications. This new version is currently in development and will be deployed as soon as testing and validation has been completed. To ensure that the issue does not reoccur before the team can deploy the fix, the Engineering and Operations teams have implemented a rate limit on the affected service so that it will not experience the unusually high volume of requests that triggered the defect.
October 9, 2018 - 7:40 PM xMatters internal monitoring alerts Operations to issue in North America region
7:47 AM Client Assistance begins testing notification delivery
8:00 PM Client Assistance escalates issue to Severity 1; incident response teams begin investigation
8:30 PM Notification service restarted.
8:40 PM Client Assistance confirms notifications are being processed; all services restored.
If you have any questions, please visit http://support.xmatters.com