/[pdpsoft]/nl.nikhef.pdp.dynsched-pbs-plugin/trunk/RELEASE
ViewVC logotype

Contents of /nl.nikhef.pdp.dynsched-pbs-plugin/trunk/RELEASE

Parent Directory Parent Directory | Revision Log Revision Log


Revision 2146 - (show annotations) (download)
Fri Jan 14 16:11:45 2011 UTC (11 years, 8 months ago) by templon
File size: 5223 byte(s)
Merge release 2.2.0 changes with trunk
1 This file contains release notes and a change history for the
2 pbs/torque and maui plugins for the lcg-info-dynamic-scheduler information provider.
3 The notes are most recent first.
4
5 Release 2.2.0
6 Moved to new repository format and changed RPM name. Rather large changes in two areas. First, in the logFileParser
7 part of the pbsServer.py code. The new "account" field printed in the torque accounting logs broke the old parser,
8 this is fixed in the current release. The second set of changes have to do with how the LiveServer code parses
9 the 'qstat -f' output. When present (recent versions of torque), the code now uses 'startTime' as the start time
10 of the job, and uses the 'euser' and 'egroup' as the user and group of the job. If these fields are not present,
11 the code reverts to the old behavior. Using startTime will give slightly different results, since startTime
12 is present immediately upon job start, whereas walltime (used in the old calculation) is only updated once per
13 mom update interval, meaning there were always a few jobs in the system that were in state "running" but that had
14 no startTime. This will no longer be the case.
15
16 Three user-visible changes: there are two new job attributes in the "Job" object. One is the walltime used by the job,
17 the other is the 'startAnchor' field which tells you whether the code has used the 'startTime' to find the start time,
18 or has done it the old way by subtracting the walltime used from the current time. These data are both printed
19 by lrmsinfo-pbs. Those two changes are in principle user-visible, but only for those users directly using lrmsinfo-pbs.
20 Finally, vomaxjobs-maui accepts an argument '-k' to provide a key-file to the diagnose command. This is needed for
21 e.g. installations where the diagnose client cmd is on RHEL4 and the maui server is on RHEL5 (different build = different key).
22
23 Information below this point refers to older versions of the package, under the older naming
24 scheme.
25
26 Release 2.1.0
27 - Remove support for specifying hostname on command line (savannah bug 35662)
28
29 Release 2.0.1
30 - fix LICENSE, otherwise no change.
31
32 Release 2.0.0
33 - uses new lrms.py, otherwise no changes.
34
35 Release 1.6.1
36 - update documentation and examples to be consistent with 2.0 release
37
38 Release 1.6.0
39
40 - some changes to build system (three targets increases aggravation)
41 - some changes to pbsServer classes to assist in debugging.
42 - some changes to vomaxjobs-maui to assist in debugging/testing;
43 also fixed various unreported bugs discovered during testing.
44 - Change mapping of pbs/torque job states in pbs classes; up til now
45 was either queued (Q) or running (any other states). Now we have:
46
47 From the qstat (torque 2.0.0p4) man page:
48
49 C - Job is completed after having run (mapped to 'done')
50 E - Job is exiting after having run. (mapped to 'running')
51 H - Job is held. (mapped to 'pending')
52 Q - job is queued, eligible to run or routed. (mapped to 'queued')
53 R - job is running. (mapped to 'running')
54 T - job is being moved to new location. (mapped to 'pending')
55 W - job is waiting for its execution time (mapped to 'queued')
56
57 Release 1.5.2:
58
59 pbs package: Fix to vomaxjobs-maui to deal with cases where there is
60 extra 'warning' output near the top of the command output from diagnose -g.
61
62 Release 1.5.1:
63
64 fix dependency problems with RPMs.
65
66 Release 1.5.0
67
68 in vomaxjobs-maui, adapt to handle MAXPROC specifications like
69 MAXPROC=soft,hard The code reports the 'hard' limit, since
70 this is relevant when the system is not full, and this is when
71 it's needed. Maui uses the soft limit on a full system, but
72 in this case the info provider will drop FreeSlots to zero as
73 soon as jobs remain in the queued state instead of executing
74 immediately.
75
76 Release 1.4.2
77
78 in pbsServer.py: included Steve Traylen's patch to deal with jobs for which the
79 uid/gid printed by 'qstat' is not listed in the in running machine's
80 pw DB. This can happen when the CE is not the same physical
81 machine as the actual LRMS server.
82
83 Estimated Response Time Info Providers (v 1.4.1)
84 ------------------------------------------------
85
86 This information provider is new in LCG 2.7.0 and is
87 contained in two RPMs, lcg-info-dynamic-scheduler-generic
88 and lcg-info-dynamic-scheduler-pbs. Sites using torque/pbs
89 as an LRMS and Maui as a scheduler are fully supported by
90 this configuration; those using other schedulers and/or
91 LRMS systems will need to provide the appropriate back-end
92 plugins.
93
94 For sites meeting the following criteria, the system should
95 work out of the box with no modifications whatsoever:
96
97 LRMS == torque
98 scheduler == maui
99 vo names == unix group names of that vo's pool accounts
100
101 Documentation on what to do if this is not the case can be
102 found in the file
103
104 lcg-info-dynamic-scheduler.txt
105
106 in the doc directory
107
108 /opt/lcg/share/doc/lcg-info-dynamic-scheduler
109
110 There is also documentation in this directory indicating
111 the requirements on the backend commands you will need to
112 provide in the case that you are using a different
113 scheduler or LRMS. Tim Bell at CERN can help for people
114 using LSF.

Properties

Name Value
svn:keywords Id URL

grid.support@nikhef.nl
ViewVC Help
Powered by ViewVC 1.1.28