Occasional Coral server restarts ...

SNF Lab Members:

Some of you may have noticed several times over the past couple of days
when you could not access Coral (either local or remote) for a period of
a few minutes. As you likely know, we are in the process of exchanging
Coral-style interlocks with new interlocks that are compatible with
Badger. When we change this information in the database, this requires
a restart of the Coral servers. When this happens, you existing Coral
sessions will die off and you will need to start a new Coral session.
When things are going smoothly, we can update interlocks on several
tools in a single restart of the servers ... and a restart of the
servers typically happens in about 20 seconds.

However, both on Monday and Tuesday, we ran into a couple of interlocks
that were giving us problems. In trying to diagnose those problems,
reset network connections, and get those interlocks working, we had to
restart the Coral servers more times than we would have liked. I think
that we restarted the servers a total of 10-20 times over those days.
While we still don't know whether these interlocks have configuration,
firmware, or hardware problems, we were clearly encountering problems on
a couple of the interlocks that we have not experienced on the other 40
or so that have been already installed.

We apologize for the inconvenience that the switching of interlocks has
caused and will hope to keep the number of server restarts to a minimum.

Let me know if you have any questions,

John