Longjobs -- Service Status Archive

For full Longjobs documentation, see: Overview | Job Scripts | Running Jobs | Checking Job Status

Fri Feb 8 2002, 3:00-5:00 pm: planned upgrades

• The longjobs master will be brought down at 3pm Friday, Feb. 8, for hardware and software upgrades. The work is expected to take about 2 hours. At this time, a software upgrade will also be applied to the execution machines. Jobs may not be added to the queue for the duration of the outage.

  Jobs already running on an execution machine will NOT be interrupted - any such machine will receive its software upgrade after the running job completes. No jobs will be deleted from the queue.

Mon Oct 22 2001, 2:00-2:20 am: planned network outage

• Beginning at 2:00am, there will be a network outage for campus backbone work which may affect the longjobs service. The work is expected to be completed within 20 minutes; during that time, you will not be able to interact with the service. Jobs which have already started should continue to run, but they may encounter problems accessing network file servers during that time. (The outage may also affect the main web servers, in which case this site would not be available; any updated information on the outage should be available from the IS Services Status Page at http://nic.mit.edu/3down.)

Sun Oct 13 2001, 6-9am: planned machine room outage

• From 6-9am, the machine room housing the longjobs machines will be shut down for electrical work; this is a major shutdown which will also make other services including email, web.mit.edu, and various Athena services unavailable. We strongly recommend waiting to submit any new jobs until after 9am. Note that we will stop the service from starting new jobs ahead of the shutdown (this has already been done for the -long queues, and will apply to the -medium queues late Friday evening), but it will still be possible to submit to the queues until the actual shutdown. The command 'qstat -l' will show the comment "Not Running: Queue not started" if your job is in this state); such a job should run as expected once the service is back up, but it is possible that other circumstances related to this outage will cause unexpected results.

  Any updated information on the outage should be available from the IS Services Status Page at http://nic.mit.edu/3down (a web site which will not be affected by this shutdown).

August 27-31, 2001: planned upgrades, completed Aug 29

• Machines will be taken offline for updates over the course of the week; running jobs should not be interrupted, but machines serving the queues will become unavailable for some periods.

Sun Apr 22 2001: planned network outages

• [posted 7pm Thu Apr 19 2001] From 10:00-10:30am, there will be short network outages during infrastructure upgrades in w92 which may affect the longjobs service. The network is expected to lose connectivity for a minute or two, once or twice within that period; during those times, you will not be able to interact with the service. Jobs which have already started should continue to run, but they may encounter problems accessing network file servers during that time. (The outage will also affect the main web servers, so this site will not be available; any updated information on the outage should be available from the IS Services Status Page at http://nic.mit.edu/3down.)

Sun Apr 1 2001: planned network outage

• [posted 6pm Fri Mar 30 2001] Beginning at noon, there will be an upgrade to network equipment in w92 which may affect the longjobs service. The work is expected to be completed within 15 minutes; during that time, you will not be able to interact with the service. Jobs which have already started should continue to run, but they may encounter problems accessing network file servers during that time. (The outage will also affect the main web servers, so this site will not be available; any updated information on the outage should be available from the IS Services Status Page at http://nic.mit.edu/3down.)

Sun Mar 18 2001: planned network outage occurred 12:00-12:18 pm

• [posted 3pm Sat Mar 17 2001] Beginning at noon, there will be an upgrade to network equipment which may affect the longjobs service. The work is expected to be completed within 20 minutes; during that time, you will not be able to interact with the service. Jobs which have already started should continue to run, but they may encounter problems accessing network file servers during that time. (The outage will also affect the main web servers, so this site will not be available; any updated information on the outage should be available from the IS Services Status Page at http://nic.mit.edu/3down.)
• [posted 1pm Sun Mar 18 2001] The previously announced network upgrade work has been completed; it lasted from noon until approximately 12:18pm. Jobs which had already started continued to run, but output should be checked for missing or inconsistent data. (If your job tried to access network file servers during the outage, the stderror file may contain messages such as "Connection timed out".)

Mar 1 2001: test began

longjobs-help@mit.edu
Last modified: Fri Jun 14 2002