1CONDOR_DAGMAN_METRICS_REPORTER(1)HTCondor ManuaClONDOR_DAGMAN_METRICS_REPORTER(1)
2
6 condor_dagman_metrics_reporter - HTCondor Manual
7 Report the statistics of a DAGMan run to a central HTTP server
9
13 condor_dagman_metrics_reporter [-s ] [-u URL] [-t maxtime] -f
14 /path/to/metrics/file
15
17 condor_dagman_metrics_reporter anonymously reports metrics from a DAG‐
18 Man workflow to a central server. The reporting of workflow metrics is
19 only enabled for DAGMan workflows run under Pegasus; metrics reporting
20 has been requested by Pegasus' funding sources: see
21 http://pegasus.isi.edu/wms/docs/latest/funding_citing_usage.php#usage_statistics
22 and
23 https://confluence.pegasus.isi.edu/display/pegasus/DAGMan+Metrics+Reporting
24 for the requirements to collect this data.
25 The data sent to the server is in JSON format. Here is an example of
27 what is sent:
28 {
30 "client":"condor_dagman",
31 "version":"8.1.0",
32 "planner":"/lfs1/devel/Pegasus/pegasus/bin/pegasus-plan",
33 "planner_version":"4.3.0cvs",
34 "type":"metrics",
35 "wf_uuid":"htcondor-test-job_dagman_metrics-A-subdag",
36 "root_wf_uuid":"htcondor-test-job_dagman_metrics-A",
37 "start_time":1375313459.603,
38 "end_time":1375313491.498,
39 "duration":31.895,
40 "exitcode":1,
41 "dagman_id":"26",
42 "parent_dagman_id":"11",
43 "rescue_dag_number":0,
44 "jobs":4,
45 "jobs_failed":1,
46 "jobs_succeeded":3,
47 "dag_jobs":0,
48 "dag_jobs_failed":0,
49 "dag_jobs_succeeded":0,
50 "total_jobs":4,
51 "total_jobs_run":4,
52 "total_job_time":0.000,
53 "dag_status":2
54 }
55 Metrics are sent only if the condor_dagman process has PEGASUS_METRICS
57 set to True in its environment, and the CONDOR_DEVELOPERS configura‐
58 tion variable does not have the value NONE.
59 Ordinarily, this program will be run by condor_dagman, and users do not
61 need to interact with it. This program uses the following environment
62 variables:
63 PEGASUS_USER_METRICS_DEFAULT_SERVER
65 The URL of the default server to which to send the data. It de‐
66 faults to http://metrics.pegasus.isi.edu/metrics. It can be
67 overridden at the command line with the -u option.
68 PEGASUS_USER_METRICS_SERVER
70 A comma separated list of URLs of servers that will receive the
71 data, in addition to the default server.
72 The -f argument specifies the metrics file to be sent to the HTTP
74 server.
75
77 -s Sleep for a random number of seconds between 1 and 10, before
78 attempting to send data. This option is used to space out the
79 reporting from any sub-DAGs when a DAG is removed.
80 -u URL Overrides setting of the environment variable PEGA‐
82 SUS_USER_METRICS_DEFAULT_SERVER. This option is unused by
83 condor_dagman; it is for testing by developers.
84 -t maxtime
86 A maximum time in seconds that defaults to 100 seconds, set‐
87 ting a limit on the amount of time this program will wait for
88 communication from the server. A setting of zero will result
89 in a single attempt per server. condor_dagman retrieves this
90 value from the DAGMAN_PEGASUS_REPORT_TIMEOUT
91 configuration variable.
92 -f metrics_file
94 The name of the file containing the metrics values to be re‐
95 ported.
96
98 condor_dagman_metrics_reporter will exit with a status value of 0
99 (zero) upon success, and it will exit with a value of 1 (one) upon
100 failure.
101
103 HTCondor Team
104
106 1990-2022, Center for High Throughput Computing, Computer Sciences De‐
107 partment, University of Wisconsin-Madison, Madison, WI, US. Licensed
108 under the Apache License, Version 2.0.
1098.8 Jun 13, 2022CONDOR_DAGMAN_METRICS_REPORTER(1)