Monitoring.adoc

Monitoring/Validation Content Sync

You may choose to use grabbit.sh to monitor your sync, or you can validate/monitor your sync by going to the following URIs:

/grabbit/job/<jobId> /grabbit/transaction/<transactionId>

A job status has the following format :

 {
       "endTime": "Timestamp",
       "exitStatus": {
           "exitCode": "Code",
           "exitDescription": "",
           "running": "true/false"
       },
       "jcrNodesWritten": "#OfNodes",
       "jobExecutionId": "JobExecutionId",
       "path": "currentPath",
       "startTime": "TimeStamp",
       "timeTaken": "TimeInMilliSeconds",
       "transactionId": "transactionId"
   }

Couple of points worth noting here: "exitCode" can have 4 states - UNKNOWN, COMPLETED, FAILED, or VALIDATION_FAILED. UNKNOWN means the job is still running. COMPLETED means that the job was completed successfully. FAILED means the job failed. VALIDATION_FAILED means the job was aborted due to client configuration; This could mean that although the configuration was valid, Grabbit refused to perform some work due to imminent introduction of unintended consequences. "jcrNodesWritten" : This indicates how many nodes are currently written (increments by 1000) "timeTaken" : This will indicate the total time taken to complete content grab for currentPath

If exitCode returns as UNKNOWN, that means the job is still running and you should check for its status again.

Sample of a real Grabbit Job status

Two loggers are predefined for Grabbit. One for Grabbit Server and the other for Grabbit Client. They are batch-server.log and batch-client.log respectively. These log files are for anything logged in com.twcable.grabbit.server.batch and com.twcable.grabbit.client.batch packages.

If you want to see what nodes are being written on the Grabbit Client, change the logging for batch-client.log above to DEBUG or TRACE.