First up, a reminder that Zimbra is being upgraded by CIS tomorrow (9/14), with an outage window of 9AM to 3PM. During that time, Zimbra will be inaccessible. Any mail sent during that window will be delivered after the downtime.
Next up, we’ve got some maintenance work on our servers that we can’t put off any longer. And, because it’s not the type of work we can do without incurring a downtime, we’re scheduling it for a weekend, specifically September 28 and 29, beginning the morning of the 28th (exact time to be scheduled).
Due to the nature of this work, the services we’re working on will be down and offline for some period, up to 12 hours. We’re doing as much of what we can do with live machines in advance, but a lot of the work requires the machines to be down so the data can be moved to its new home. We’re still working out the likely outage windows, but what follows is the preliminary list of affected servers and services. If this presents an undue hardship to you, let me know as soon as possible so we can look at options.
Affected servers (services)
- silver.mcs.anl.gov (Mac server, administrative file shares)
- variant.mcs.anl.gov (Subversion/Trac for MCS, ALCF, KBase, MPICH)
- caveat.mcs.anl.gov (trac.mpich.org)
- coredb-2.mcs.anl.gov (backup mysql server)
- blizzard.mcs.anl.gov (collab.mcs.anl.gov aka Confluence)
- licman3.mcs.anl.gov (License server)
- dust.mcs.anl.gov (ALCF Drupal staging server)
- squall.mcs.anl.gov (ALCF Drupal production server — www.alcf.anl.gov)
- owney.mcs.anl.gov (All mailman lists, see below)
- rt.mcs.anl.gov (ticketing system) *5 minute window for reboot only
During the mailman server outage, no mail will be lost or bounced, simply delayed during the downtime. Mailing list domains that will be offline:
We expect the outages to be contained to Saturday the 28th, with the 29th reserved in case things go unexpectedly poorly.
Again, if any of this presents an undue hardship, let me know ASAP so we can explore our options for minimizing it. Thanks!