Our health-check should access both Redis and Cassandra so that a failure to connect to one of these services will force the worker in question to restart.
<rdar://problem/64118077>
One of our instances of the results database was brought down by this bug over the weekend (although not the one hosted at results.webkit.org)
Created attachment 401595 [details] Patch
Created attachment 401602 [details] Patch
Committed r262876: <https://trac.webkit.org/changeset/262876> All reviewed patches have been landed. Closing bug and clearing flags on attachment 401602 [details].