Context Navigation

Timeline

13:28 Changeset [1697] by buettner: forgot to commit these earlier
11:27 Ticket #393 (setres.py get_partitions problem) closed by buettner: fixed: r1696 ought to fix this
11:27 Changeset [1696] by buettner: fixing #393 ; setres should stop showing tracebacks (at least some of …
11:18 Ticket #422 (prologue execution in cluster_system needs a timeout) created by buettner: the network connections used to execute the prologue can and do hang make …

14:18 Ticket #421 (change nodelist ordering) closed by buettner: fixed: r1695 takes care of this
14:17 Changeset [1695] by buettner: implementing #421 ; using the order of the hostfile to determine the order …

16:35 Ticket #421 (change nodelist ordering) created by buettner: nodelist shows the cluster nodes in "alphabetic" order, which means vs11 …
16:14 Changeset [1694] by hnaik: added ftb-standalone-test.py
16:13 Changeset [1693] by hnaik: Cleanup
10:57 Ticket #414 (qsub -h flag results in log file owned by root) closed by buettner: fixed: this should be resolved by r1673
10:54 Ticket #347 (queue transitivity issue) closed by buettner: fixed: i think this isn't really an issue now, with the various changes to …
10:53 Ticket #346 ([alcf-internal #16616] handle scheduling over asymmetric partitions ...) closed by buettner: fixed: r1469 first addressed this, and several other revisions have as well. I …
10:49 Ticket #352 (cobalt for clusters is leaving nodes in the allocated state when they ...) closed by buettner: fixed: duplicate of #420
10:48 Ticket #291 ([alcf-internal #7575] -free partition at end of script job) closed by buettner: fixed: r1681 should take care of this
10:43 Ticket #394 (cluster_system trying to act on partitions) closed by buettner: fixed: r1692 which implemented #420 should also fix this
10:41 Ticket #420 (Cluster nodes not deallocated if qdel used too late) closed by buettner: fixed: r1692 should fix this
10:41 Changeset [1692] by buettner: fixing #420 ; found the place to deal with a hasty job kill

17:57 Changeset [1691] by hnaik: weird error with using FEA instance for publish
17:40 Changeset [1690] by hnaik: Something wrong with calling of Declare_pub_events
14:17 Ticket #420 (Cluster nodes not deallocated if qdel used too late) created by tstacey: On the clusters, if you qdel a job after Cobalt starts trying to run it …
13:56 Changeset [1689] by buettner: some minor changes to the code for #279 ; changing logging and reducing …

17:22 Changeset [1688] by hnaik: Policy Interface added
17:16 Ticket #279 (Would like an option for bgsystem to refresh its list of partitions from ...) closed by buettner: fixed: r1687 implements this, although the refreshing is continuous rather than …
17:14 Changeset [1687] by buettner: implementing #279 ; although rather than requiring signalling from an …
12:57 Changeset [1686] by hnaik: Policy Configuration Interface

10:06 Changeset [1685] by buettner: version bump to 0.99.0pre7
09:54 Changeset [1684] by buettner: tagged 0_99_0pre7 release
09:54 Changeset [1683] by buettner: redoing 0_99_0pre7 release?
09:49 Changeset [1682] by buettner: tagged 0_99_0pre7 release
05:19 Changeset [1681] by toonen: When cobalt job completes, the partition the job assigned and the all …
05:02 Changeset [1680] by toonen: added a svn:ignore property for src/lib/DataTypes

09:14 Ticket #419 (showres output "funny") closed by buettner: fixed: r1678 fixes this
09:13 Changeset [1678] by buettner: fixing #419 ; mergelist has been rewritten and moved into Cobalt.Util
08:55 Ticket #419 (showres output "funny") created by buettner: on eureka, showres was bungling the display of a "condensed" list of nodes …

16:52 Ticket #418 (cluster simulator is broken) closed by buettner: fixed: r1677 fixes this
16:51 Changeset [1677] by buettner: fixing #418 ; the cluster simulator didn't get the new ProcessGroup? stuff
16:36 Ticket #418 (cluster simulator is broken) created by buettner: the cluster simulator can't start

15:14 Ticket #417 (Propagate queue deletion to partitions) created by cheetah: When a partition is deleted with cqadm, an option should exist to …

15:04 Ticket #413 (draining fails) closed by buettner: fixed: r1676 should fix this
15:01 Changeset [1676] by buettner: fixing #413 ; making sure that drain choices immediately impact other jobs …

18:43 Ticket #416 (cobalt-mpirun fails on certain -env strings) created by rloy: Usually it works to give multiple variable settings in a single -env …
17:16 Changeset [1675] by buettner: fixing broken test file
17:02 Changeset [1674] by toonen: added a method to destroy (free) a partition. updated the methods that …
16:55 Changeset [1673] by toonen: modified Job.write_cobalt_log so that it will not create the cobaltlog …
14:54 Ticket #415 (traceback) closed by buettner: fixed: r1672 fixes this
14:53 Changeset [1672] by buettner: fixing #415 ; traceback from uninitialized variable
14:51 Ticket #415 (traceback) created by buettner: local variable 'response' referenced before assignment Traceback (most …

14:33 Changeset [1671] by buettner: as part of #409 make sure to handle the case that the stdin file cannot be …
09:53 Ticket #409 (need workaround for cobalt-mpirun redirection) closed by buettner: fixed: r1670 implements this (the flag based version)
09:49 Changeset [1670] by buettner: implementing feature #409 ; letting the user specify --stdin, --stdout, …

17:57 Ticket #414 (qsub -h flag results in log file owned by root) created by cheetah: When submitting a job with -h (user hold), <jobid>.cobaltlog ends up owned …
16:54 Ticket #413 (draining fails) created by buettner: over the past week, i noticed that when "queued" jobs were in the special …

Note: See TracTimeline for information about the timeline view.