On Oracle SOA Suite 12c (and technically Oracle WebLogic Server 12c), I noticed that a couple of managed servers were in a RUNNING state but the health was reporting Failed.

What could cause this particular scenario?

In this domain, the TLOGs were persisted in the database and the Persistent Store used to store the TLOGS returned a SetPrimaryStore failed exception. This was because the data source was suspended (e.g., Pool wls_tlogs is Suspended).

There apparently was a temporary hiccup in the database that caused this, and thus WebLogic couldn't write to the TLOG, hence the Failed health status.