How to create a workload profile
To profile the NetBackup workload, you can use the bpdbjobs command to collect the job detail report. The following is the command syntax to collect the job report:
/usr/openv/netbackup/bin/admincmd/bpdbjobs -report -all_columns > /tmp/bpdbjobs.out
Note:
To get a job summary, change the -all_columns option to -most_columns.
In the bpdbjobs report, many fields are included in each line of the job record. The following fields are useful to analyze the policy and job information.
Field | Description |
|---|---|
field2 | Job type (for example, backup, restore, duplicate, replication, and so on) |
field3 | State of the job (for example, queued, active, done, and so on) |
field8 | Media server that is used by the job |
field9 | Job started time |
field10 | Elapsed time for the job |
field11 | Job end time |
field22 | Policy type (for example Standard, Oracle, NDMP, VMware, and so on) |
field58 | Deduplication rate |
field61 | Deduplication ratio percent |
More details of fields for the bpdbjobs command can be found in NetBackup Commands Reference Guide
Depending on the workload analysis target, the corresponding field can be used to analyze the jobs.
The following shows an example command to count the number of jobs for each type.
# awk -F, '{print $2}' /tmp/bpdbjobs.out | sort | uniq -c
7550 0 -> backup
9188 20 -> replication
8932 28 -> snapshot