Video Screencast Help

How to measure throughput when backing up

Created: 13 Nov 2012 • Updated: 20 Jan 2013 | 1 comment
This issue has been solved. See solution.

Hello Everybody!!

I would like to know how can I measure the throughput from backups because I have gotten very low throughput when backing up. Some times the backups end with a good time and some times we have to wait up to 58 hours when it normally run during  8 hours.

I want to tanks any advaice regarding this issue.

Also I am looking for a bandwith issue and also the load from the server on that period of time.

Best regards 

Comments 1 CommentJump to latest comment

Yasuhisa Ishikawa's picture

Could you tellus more detail about your environment like version of NetBackup and OS, type of backup,...?
For file backup, bpbkar debug log with level 5 is useful to understand which directories and files it spent time to backup.

Can you find any regularity? Significant performance degration sometimes come from maintenance operations on storage like parity check. Degration also occurs when DNLC is cleared somehow.

Observing "wait and delay counter" is good starting point to isolate your problem. You can determine whether the bottleneck resides on, media server or 'network to client' side with it. Please check 'Backup planning and performance tuning guide" for more detail.

Also check disk activity and busy rate on client using iostat(UNIX) or perfmon(Windows)

Authorized Symantec Consultant(ASC) Data Protection in Tokyo, Japan