vocms0310
Agent tweaks
- AgentStatusWatcher disabled ==> OK
- Set maxRetries to 0 ==> OK
- Restart the necessary components ==> OK
$manage execute-agent wmcoreD --restart --components=RetryManager,ErrorHandler,AgentStatusWatcher
- Enable ALL sites and set a high number of thresholds for them ==> OK
$manage db-prompt wmagent
UPDATE wmbs_location SET state=(SELECT id from wmbs_location_state where name='Normal') WHERE state!=(SELECT id from wmbs_location_state where name='Normal');
UPDATE wmbs_location SET running_slots=2000, pending_slots=1000;
UPDATE rc_threshold SET max_slots=2000, pending_slots=1000;
Draining logs
cmst1@vocms0310:/data/srv/wmagent/current $ python drainAgent.py
*** Amount of jobs in condor per workflow, sorted by condor job status:
{}
*** WORKFLOWS: found 13 distinct workflows in this agent.
prozober_ACDC0_SUS-RunIISpring16FSPremix-00083_00047_v0__170130_100804_7952 completed
pdmvserv_task_JME-PhaseIFall16GS-00003__v1_T_170127_092523_4740 completed
pdmvserv_task_BPH-PhaseIIFall16GS82-00011__v1_T_170124_203656_2417 announced
prozober_ACDC0_SUS-RunIISpring16FSPremix-00086_00047_v0__170130_104559_455 completed
pdmvserv_task_HIG-RunIISummer16DR80Premix-01911__v1_T_170116_133411_9741 completed
pdmvserv_task_SUS-RunIISummer16DR80Premix-00135__v1_T_161124_210645_3749 announced
vlimant_task_JME-RunIISummer16DR80-00010__v1_T_161128_025525_6414 aborted-archived
pdmvserv_task_TOP-RunIISummer16DR80Premix-00083__v1_T_161219_123711_1390 running-closed
pdmvserv_task_HIG-RunIISummer16DR80Premix-01880__v1_T_170113_162606_6082 running-closed
pdmvserv_SUS-RunIISpring16FSPremix-00083_00047_v0__170116_081703_7893 completed
wmagent_dmason_TEST_GOOGLE_161110_204157_8813 aborted-archived
pdmvserv_task_EXO-RunIISummer15wmLHEGS-04239__v1_T_170114_004935_4431 announced
wmagent_dmason_TEST_GOOGLE_161111_134234_5412 aborted-archived
*** WORKFLOWS: there are 1 distinct workflows not completed.
wmagent_dmason_TEST_GOOGLE_161110_204157_8813 aborted-archived
*** WORKFLOWS: found 0 workflows not fully injected.
*** WMBS: amount of wmbs jobs in each status:
[{'count': 105384, 'name': 'cleanout'}, {'count': 11, 'name': 'executing'}]
*** WMBS: 2 workflows with executing jobs in wmbs:
wmagent_dmason_TEST_GOOGLE_161110_204157_8813 aborted-archived
wmagent_dmason_TEST_GOOGLE_161111_134234_5412 aborted-archived
*** SUBSCRIPTIONS: subscriptions not finished: 3
vlimant_task_JME-RunIISummer16DR80-00010__v1_T_161128_025525_6414 aborted-archived
wmagent_dmason_TEST_GOOGLE_161110_204157_8813 aborted-archived
wmagent_dmason_TEST_GOOGLE_161111_134234_5412 aborted-archived
*** SUBSCRIPTIONS: found 4 files available in WMBS (waiting for job creation):
[{'count(*)': 39, 'subscription': 1491}, {'count(*)': 83, 'subscription': 2181}, {'count(*)': 5, 'subscription': 20733}, {'count(*)': 5, 'subscription': 1490}]
*** SUBSCRIPTIONS: found 0 files acquired in WMBS (waiting for jobs to finish):
[]
*** DBS: found 0 blocks open in DBS. Printing the first 20 blocks only:
[]
*** DBS: found 27 files not uploaded to DBS.
==> Which maps to 1 unique datasets:
set(['/Muplus_Pt10-gun/RunIISummer16DR80Premix-PUMoriond17_80X_mcRun2_asymptotic_2016_TrancheIV_v6-v1/GEN-SIM-RAW'])
... that were NOT produced by any agent-known workflow. OR, the wfs are gone already.
*** PHEDEX: found 0 files not injected in PhEDEx, with valid block id (recoverable).
*** PHEDEX: found 481745 files not injected in PhEDEx, with valid block id (unrecoverable).
==> Which maps to 3229 unique datasets:
...tons of datasets...
'/ttZJets_13TeV_madgraphMLM/RunIISummer15wmLHEGS-MCRUN2_71_V1-v1/GEN-SIM'])
... that were NOT produced by any agent-known workflow. OR, the wfs are gone already.
I'm done!
Cleaning these workflows that are - in theory - long gone from the agent. Those 3 workflows showing up with subscriptions not finished
cmst1@vocms0310:/data/srv/wmagent/current $ $manage execute-agent kill-workflow-in-agent vlimant_task_JME-RunIISummer16DR80-00010__v1_T_161128_025525_6414
Executing kill-workflow-in-agent vlimant_task_JME-RunIISummer16DR80-00010__v1_T_161128_025525_6414 ...
Canceling work for workflow: set(['vlimant_task_JME-RunIISummer16DR80-00010__v1_T_161128_025525_6414'])
Aborted worqueue elements:
[]
cmst1@vocms0310:/data/srv/wmagent/current $
cmst1@vocms0310:/data/srv/wmagent/current $ $manage execute-agent kill-workflow-in-agent wmagent_dmason_TEST_GOOGLE_161110_204157_8813
Executing kill-workflow-in-agent wmagent_dmason_TEST_GOOGLE_161110_204157_8813 ...
Canceling work for workflow: set(['wmagent_dmason_TEST_GOOGLE_161110_204157_8813'])
Aborted worqueue elements:
[]
cmst1@vocms0310:/data/srv/wmagent/current $
cmst1@vocms0310:/data/srv/wmagent/current $ $manage execute-agent kill-workflow-in-agent wmagent_dmason_TEST_GOOGLE_161111_134234_5412
Executing kill-workflow-in-agent wmagent_dmason_TEST_GOOGLE_161111_134234_5412 ...
Canceling work for workflow: set(['wmagent_dmason_TEST_GOOGLE_161111_134234_5412'])
Aborted worqueue elements:
[]
cmst1@vocms0310:/data/srv/wmagent/current $
these workflows should be gone from the output of drainAgent.py in the next time we run it. CHECK
Draining logs - take 2
cmst1@vocms0310:/data/srv/wmagent/current $ python drainAgent.py
*** Amount of jobs in condor per workflow, sorted by condor job status:
{}
*** WORKFLOWS: found 10 distinct workflows in this agent.
prozober_ACDC0_SUS-RunIISpring16FSPremix-00083_00047_v0__170130_100804_7952 completed
pdmvserv_task_JME-PhaseIFall16GS-00003__v1_T_170127_092523_4740 completed
pdmvserv_task_BPH-PhaseIIFall16GS82-00011__v1_T_170124_203656_2417 announced
prozober_ACDC0_SUS-RunIISpring16FSPremix-00086_00047_v0__170130_104559_455 completed
pdmvserv_task_HIG-RunIISummer16DR80Premix-01911__v1_T_170116_133411_9741 completed
pdmvserv_task_SUS-RunIISummer16DR80Premix-00135__v1_T_161124_210645_3749 announced
pdmvserv_task_TOP-RunIISummer16DR80Premix-00083__v1_T_161219_123711_1390 running-closed
pdmvserv_task_HIG-RunIISummer16DR80Premix-01880__v1_T_170113_162606_6082 running-closed
pdmvserv_SUS-RunIISpring16FSPremix-00083_00047_v0__170116_081703_7893 completed
pdmvserv_task_EXO-RunIISummer15wmLHEGS-04239__v1_T_170114_004935_4431 announced
*** WORKFLOWS: there are 0 distinct workflows not completed.
*** WORKFLOWS: found 0 workflows not fully injected.
*** WMBS: amount of wmbs jobs in each status:
[{'count': 91920, 'name': 'cleanout'}]
*** SUBSCRIPTIONS: subscriptions not finished: 0
*** SUBSCRIPTIONS: found 0 files available in WMBS (waiting for job creation):
[]
*** SUBSCRIPTIONS: found 0 files acquired in WMBS (waiting for jobs to finish):
[]
*** DBS: found 0 blocks open in DBS. Printing the first 20 blocks only:
[]
*** DBS: found 27 files not uploaded to DBS.
==> Which maps to 1 unique datasets:
set(['/Muplus_Pt10-gun/RunIISummer16DR80Premix-PUMoriond17_80X_mcRun2_asymptotic_2016_TrancheIV_v6-v1/GEN-SIM-RAW'])
... that were NOT produced by any agent-known workflow. OR, the wfs are gone already.
*** PHEDEX: found 0 files not injected in PhEDEx, with valid block id (recoverable).
*** PHEDEX: found 481745 files not injected in PhEDEx, with valid block id (unrecoverable).
==> Which maps to 3229 unique datasets:
...
... that were NOT produced by any agent-known workflow. OR, the wfs are gone already.
I'm done!
and the agent is READY for redeployment.