[Halld-offline] uconn srm upgrade

[Richard Jones]

Over the weekend, I noticed that a rising percentage (now 25%) of our
production jobs on the osg were failing when trying to save their results
back to the uconn srm. What has happened is that OSG production managers
have advised sites to upgrade all of their server certs, after patching
their systems for the heart-bleed security hole in openssl.  At the end of
last week, a large number of sites did so.  Our UConn site was not
vulnerable to Heartbleed because of a fluke in our setup, but we had to
upgrade because our server certs all timed out last week.

The Northwestern srm is working, but Sean does not believe it has been
tested thoroughly enough to put into production for a big workload like
this, so I have decided to halt production this morning, take down our srm,
and upgrade it.

Sorry for the inconvenience to any of you who were trying to copy files
over.  The UConn srm is going down at 9:00am today and coming back as soon
as possible, hopefully before 2:00 this afternoon.

-Richard J.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20140421/d6a82102/attachment-0002.html>