Skip to main content

 

Druva Documentation

Job behavior

This topic consists of the following sections:

Skipped job

The scheduled job that did not start during the scheduled window due to another running job, remains in the queue until the window is active, and after the window expiry, the job is marked as Skipped. The job is marked as skipped in the following scenarios:

  • If the first running job is a manually-triggered backup, that is Backup Now that runs with no backup window and the subsequent scheduled job is not picked by the client.
  • If the first job is a scheduled job with no backup window defined, that is when the Backup Duration is '0', and the subsequent scheduled job is not picked by the client.
  • If the first job is a scheduled job with a very high backup window, for example, 96 hours, and the subsequent job is not picked by the client.
    Please refer error code PHOENIX 230.

Disconnected client

  • For a Backup Now job, if the client is disconnected when the job is triggered, the backup job fails as follows: 
    • For File server and MS-SQL server: The backup job fails immediately with the PHOENIX 17 error.
    • For a virtual machine with proxy load balancer disabled: The backup job fails immediately with the PHOENIX 17 error.
      The proxy load balancer is disabled if one or more backup proxy client in your registered vCenter are below version 4.6.2.
    • For a virtual machine with load balancer enabled: The backup job fails after 24 hours.
      The proxy load balancer is enabled if all the backup proxy clients in your registered vCenter to version 4.6.2 and later.

      Note: For VMware with load balancer enabled, the job may not be picked by client as all backup proxies in the backup proxy pool may be disconnected or the backup proxy pool is busy. In this case, PHOENIX 229 error message is displayed.

  • For scheduled jobs, if the job did not reach the client as the client was disconnected throughout the job window. 
    • For File server and MS-SQL server: The backup job will fail after the scheduled window expires with PHOENIX 228.
    • For a virtual machine with load balancer disabled: The backup job will fail after the expiry of the scheduled window with PHOENIX 227.
      The proxy load balancer is disabled if one or more backup proxy client in your registered vCenter are below version 4.6.2.
    • For a virtual machine with load balancer enabled: The backup job will fail after the expiry of the scheduled window with PHOENIX 231.
      Proxy load Balancer is enabled if all the backup proxy client in your registered vCenter to version 4.6.2 and later.

Note: After triggering the schedule job with infinite window, if the client is disconnected for 24 hours, the job fails.

Backup proxy pool busy

  • For a Backup Now job, if the backup proxy pool is busy, the job will fail with the following error:

    VMware with Proxy Load balancer enabled: Fail the backup job after 24 hours with PHOENIX 229 error.
     
  • For a Scheduled Backup, if the backup proxy pool is busy during the scheduled window, the job will fail with the following error:

    VMware with Proxy Load balancer enabled: Fail the backup after backup window expires with PHOENIX 231 error.
     
  • If the job is in the client queue but was not completed, the job will fail with PHOENIX 158 error.

First backup behavior

Typically, the first backup job for a server requires a lot of time because the data set is large. For a large data set getting backed up, it is possible that the backup duration you specify in the backup policy is not enough for the job to get completed. A backup job is considered complete when a restore point is created at the end of the job. To ensure that the first backup job gets completed successfully, the Ignore backup duration for first backup option is available in the backup policy. By default, this option is enabled to ignore backup duration until the first restore point is created, and the server is backed up for the first time. 

While the first backup job is running and the Ignore backup duration for first backup option is selected:

  • Specified backup duration in backup policy is ignored
  • Any subsequent scheduled or manual backup jobs triggered during that duration are not triggered in order to ensure the successful run of the first job.

Until the first restore point is created, Phoenix displays the following message if you trigger a manual backup job during an ongoing backup.

After a restore point is created, all subsequent backup jobs are treated as incremental backups and backup policy settings are applied.

New backup window expired state for jobs

A backup job gets the Backup Window Expired state if there is some data that is backed up but the restore point is not created because backup duration specified in the backup policy is not enough to get the entire data set backed up. 

If many jobs end with the Backup Window Expired state, review the backup duration specified in backup policy because it is not enough for the backup to complete. However, a user can ignore this message because the entire data set is eventually backed up when subsequent scheduled jobs are triggered as Phoenix always performs an ever incremental backup. 

Job statistics for backup and restore jobs

Phoenix provides additional job statistics for backup and restores jobs. The job statistics serve as instrumentation bits for troubleshooting backup performance issues.

The job stats are listed on the Job Details page. The following table captures the behavior of the old, new, and upgraded clients when old or new backup and restore jobs are run.

Client vs. Job Old Client New Client Upgraded Client

Old Backup Job

  • The Summary tab displays the old backup job stats.
  • The Summary tab displays the renamed labels of job stats, as listed follows:
    • Estimated Data -> Source Data Scanned
    • Bandwidth Used -> Bandwidth Consumed
  • Upgrade agent to the latest version to view more backup statistics.

NA

  • The Summary tab displays the old backup job stats.
  • The Summary tab displays the renamed labels of job stats, as listed follows:
    • Estimated Data -> Source Data Scanned
    • Bandwidth Used -> Bandwidth Consumed

New Backup Job

  • The Summary tab displays the old backup job stats.
  • The Summary tab displays the renamed labels of job stats, as listed follows:
    • Estimated Data -> Source Data Scanned
    • Bandwidth Used -> Bandwidth Consumed
  • Upgrade agent to the latest version to view more backup statistics.

Summary tab displays the new backup job stats. For more information about these stats, see Job Filters.

Summary tab displays the new backup job stats. For more information about these stats, see Job Filters.

Old Restore Job

  • The Summary tab displays the old restore job stats.
  • The Summary tab displays the renamed labels of job stats, as listed follows:
    • Estimated Data -> Restored Data
  • Upgrade agent to the latest version to view more restore statistics.

NA

  • Summary tab displays the old restore job stats.
  • Summary tab displays the renamed labels of job stats, as listed follows:
    • Estimated Data -> Restored Data

New Restore Job

  • The Summary tab displays the old restore job stats.
  • The Summary tab displays the renamed labels of job stats, as listed follows:
    • Estimated Data -> Restored Data
  • Upgrade agent to the latest version to view more restore statistics.

Summary tab displays the new restore job stats. For more information about these stats, see Job Filters.

Summary tab displays the new restore job stats. For more information about these stats, see Job Filters.

Backup job statistics

The following table lists the backup job statistics specific to workloads, such as VMware, File server, MS-SQL server, and so on.

Job Stats Windows File Server Linux File Server UNC Share MS-SQL Server VMware
Status = Running 
Source Data Scanned Yes Yes Yes Yes Yes
# VMDK(s) No No No No Yes
# Files Yes Yes Yes No No
File Size Distribution Yes Yes Yes No No
# DBs No No No Yes No
Estimated Duration Yes Yes Yes Yes Yes
Data Transferred to Cloud Yes Yes Yes Yes Yes
Data Transferred to CloudCache Yes Yes Yes Yes Yes
Backup Duration No No No No No
Backup Speed No No No No No
Bandwidth Consumed No No No No No
Disk Read Rate Yes Yes Yes Yes Yes
# Network Retries Yes Yes Yes Yes Yes
Network Retry Duration Yes Yes Yes Yes Yes
CBT No No No No Yes
Pre Script Status Yes Yes No No No
Post Script Status Yes Yes No No No
Status = Failed/Successful/Successful with errors
Source Data Scanned Yes Yes Yes Yes Yes
# VMDK(s) No No No No Yes
# Files Yes Yes Yes No No
File Size Distribution Yes Yes Yes No No
# DBs No No No Yes No
Estimated Duration Yes Yes Yes Yes Yes
Data Transferred to Cloud Yes Yes Yes Yes Yes
Data Transferred to CloudCache Yes Yes Yes Yes Yes
Backup Duration Yes Yes Yes Yes Yes
Backup Speed Yes Yes Yes Yes Yes
Bandwidth Consumed Yes Yes Yes Yes Yes
Disk Read Rate Yes Yes Yes Yes Yes
# Network Retries Yes Yes Yes Yes Yes
Network Retry Duration Yes Yes Yes Yes Yes
CBT No No No No Yes

Restore job statistics

The following table lists the restore job statistics specific to the workload, such as VMware, File server, MS-SQL server, and so on.

Job Stats Windows File Server Linux File Server UNC Share MS-SQL Server VMware
Destination Location Yes Yes Yes Yes Yes
Restored Data Yes Yes Yes Yes Yes
File Size Distribution Yes Yes Yes No No
Restore Speed Yes Yes Yes Yes Yes
Disk Write Rate Yes Yes Yes Yes Yes
# Network Retries Yes Yes Yes Yes Yes
Network Retry Duration Yes Yes Yes Yes Yes