[intrepid-notify] Additional maintenance downtime tomorrow
William (Bill) E. Allcock
allcock at alcf.anl.gov
Thu Jan 29 12:44:55 CST 2009
All,
The ALCF support team and Myricom continue to work as hard as we can
to resolve the issues in our Myricom switch complex.
On Monday and Tuesday we ran diagnostic tests and identified over 100
ports that exhibit characteristics indicative of the port failures we
have been seeing. These failures present themselves in the network as
excessive CRC errors and often result in file system and login
problems until we can track them down and disable them. We pushed
Myricom as hard as we could to replace those ports today during our
already extended, but pre-scheduled, maintenance window. They
informed us that the soonest they could be ready was tomorrow, and
that they needed a 12 hour window to complete their work.
As a result, we will be putting a reservation in place for tomorrow
from 8AM - 8PM. We apologize for this additional loss of science
time. We are acutely aware of the fact that some of you have on-going
time sensitive campaigns. However, an additional 12 hours tomorrow
will allow us to get these current problem ports out of the system,
and, hopefully, buy us improved stability while Myricom works out root
cause and gets us a long term solution. Please contact support at alcf.anl.gov
with any questions, comments, or concerns.
Thank you for your continued patience,
The ALCF Support Team
More information about the intrepid-notify
mailing list