Job runs Ok manually, butt fails scheduled

[Xavier Humbert]

Hi,
Since upgrade in 5.17, one of my users has trouble with all their jobs.
In short they run manually, but not scheduled.

Here is a job which is impacted :

- defaultTab: nodes
  description: purge des logs des jobs dans /var/log/outils
  executionEnabled: true
  id: a229ef45-ff46-4908-80c5-e4ce3f8a4ca1
  loglevel: INFO
  name: Purge logs rundeck
  nodeFilterEditable: false
  nodefilters:
    dispatch:
      excludePrecedence: true
      keepgoing: true
      rankOrder: ascending
      successOnEmptyNodeFilter: false
      threadcount: '1'
    filter: 'hostname:
"pr-ofi-bddsrep01.dsofint.in.phm.education.gouv.fr" hostname:
      "pr-ofi-bddsrem01.dsofint.in.phm.education.gouv.fr" hostname:
"pr-ofi-bddsnom01.dsofint.in.phm.education.gouv.fr"'
  nodesSelectedByDefault: true
  notification:
    onfailure:
      email:
        attachLog: true
        attachLogInline: true
        recipients: l.ce...@vlist.education.gouv.fr
        subject: '[SYGNE] [PP] RunDeck - Job ${job.name} en ERREUR '
  notifyAvgDurationThreshold: null
  plugins:
    ExecutionLifecycle: {}
  schedule:
    month: '*'
    time:
      hour: '10'
      minute: '10'
      seconds: '0'
    weekday:
      day: '3'
    year: '*'
  scheduleEnabled: true
  sequence:
    commands:
    - script: |-
        #!/bin/bash
        #purge des logs de + de 30 jours dans /var/log/outils

        set -uo pipefail

        #creation fichier de log
        LOGFILE=/var/log/outils/$(basename $0|cut -d. -f1).$(date
+%Y%m%d%H%M%S).log
        touch $LOGFILE
        date>$LOGFILE
        CODERET=0

        #Main
        echo "suppression des logs de /var/log/outils de + de 30
jours"|& tee -a $LOGFILE
        sudo find /var/log/outils/ -type f -name '*.log' -mtime +30
-exec rm -f {} \; |& tee -a $LOGFILE
        CODERET=$?

        exit $CODERET
    keepgoing: false
    strategy: node-first
  user: mboucaud
  uuid: a229ef45-ff46-4908-80c5-e4ce3f8a4ca1

When run manually, everything is OK (I verified it), but when run
scheduled, he gets this error :


com.dtolabs.rundeck.core.NodesetEmptyException: No matched nodes:
NodeSet{includes={hostname=pr-ofi-bddsrep01.dsofint.in.phm.education.gouv.fr,pr-ofi-bddsrem01.dsofint.in.phm.education.gouv.fr,pr-ofi-bddsnom01.dsofint.in.phm.education.gouv.fr,
dominant=false, }}
    at
com.dtolabs.rundeck.core.execution.workflow.BaseWorkflowExecutor.validateNodeSet(BaseWorkflowExecutor.java:880)
    at
com.dtolabs.rundeck.core.execution.workflow.NodeFirstWorkflowExecutor.executeWorkflowImpl(NodeFirstWorkflowExecutor.java:92)
    at
com.dtolabs.rundeck.core.execution.workflow.BaseWorkflowExecutor.executeWorkflow(BaseWorkflowExecutor.java:220)
    at
com.dtolabs.rundeck.core.execution.WorkflowExecutionServiceThread.runWorkflow(WorkflowExecutionServiceThread.java:95)
    at
com.dtolabs.rundeck.core.logging.LoggingManagerImpl$MyPluginLoggingManager.runWith(LoggingManagerImpl.java:146)
    at
com.dtolabs.rundeck.core.execution.WorkflowExecutionServiceThread.run(WorkflowExecutionServiceThread.java:77)
Exception: class com.dtolabs.rundeck.core.NodesetEmptyException: No
matched nodes:
NodeSet{includes={hostname=pr-ofi-bddsrep01.dsofint.in.phm.education.gouv.fr,pr-ofi-bddsrem01.dsofint.in.phm.education.gouv.fr,pr-ofi-bddsnom01.dsofint.in.phm.education.gouv.fr,
dominant=false, }}
No matched nodes:
NodeSet{includes={hostname=pr-ofi-bddsrep01.dsofint.in.phm.education.gouv.fr,pr-ofi-bddsrem01.dsofint.in.phm.education.gouv.fr,pr-ofi-bddsnom01.dsofint.in.phm.education.gouv.fr,
dominant=false, }}

I doubled checked also that the nodes are correctly defines in
resources.json

I must also admit that I was unable to reproduce the problem : I
duplicated this job, the duplicate runs fine scheduled.

Regards,

Xavier

--
Xavier Humbert
CRT Supervision et Exploitation de Niveau 1
Direction Régionale Académique des Systèmes d’Information Grand Est
Rectorat de l'Académie de Nancy-Metz
Ministère de l'Éducation Nationale, de l'Enseigenement Supérieur et de la Recherche
03 83 86 27 39