This issue was reported by users who were unable to login to Dr. Know products.
After a prompt technical review and communication with customers, we learned that the issue was impacting some users, some of the time. Many users were not experiencing the issue. Further, several users who were impacted, we able to eventually login after several attempts, suggesting inconsistent or intermittent behavior.
This issue resolved itself when the Dr. Know API was re-published, without a patch. This suggests that the API had gotten into an unexpected state and needed a reboot.
The underlying issue was related to communication with an internal Azure Redis Cache instance. However, there were no errors or issues reported by the cache itself. Rather, the communication between the API and cache was not working as expected.
There were no Dr. Know releases, system updates, or patches that could have caused this incident.
We likely could have performed a restart operation on the Azure API instance, but instead, did this via a new publish, bumping up the patch version number (v.2.30.3). This immediately resolved the issue.
In the future, we may want to enhance logging around this Redis cache access to provide additional information as the issue was not able to be reproduced in a development setting - even talking to the same Redis cache instance.
Another possible cause could be an internal Microsoft Azure update?