We had a problem in one of the Database clusters. Services got affected for customers on that cluster.
Time : Aug 26th, 8:16 pm - 8:20 pm (PDT), 8:23 pm - 8:58 pm (PDT).
While
majority of the customers were unaffected, few APMInsight customers who
were on that cluster, would see their agents in down state.
Reason
: Since the user information was not accessible from the db server,
the agents were marked as deleted. We have fixed and handled it in a
different way and updated our build too.
To get back the agents in working mode, please delete the agent info file, and then restart the server.