Factors affecting the NDMP backups

[tejas9024]

Hello,

In our environment, I come across situations where NDMP backup of some shares (with larger size) completes before the shares of smaller size.

I would like to know what are the factors which lead to long backup times for NDMP backups.

I do know that long backup durations are seen for shares containing millions of files. Are there any other parameters which will lead to increase in backup times or is this the only one?

[RamNagalla]

you can try configuring the Snaphost of the filer and take the backup of the snapshots..

you can configured the nightly snapshots in Filer and then take the backup of those snapshots using the Netbackup.

if the filer is in supported device, you can look for the option Replication Director in the netbackup.

http://www.symantec.com/tv/products/details.jsp?vid=1483740597001

[jim_dalton]

All other things being equal, then yes the cifs data profile affects the backup. Each file affects the performance thus the fewer files the "better". NDMP has to traverse the tree to figure out whats what: you can see these steps in the ndmp backup jobs to give you an idea. Anything else when things arent so equal...? Well yes: the number of users accessing the filer, any other processes eg virus scanner, the aggregate, any other users of the aggregate, disk numbers in the agg, disk types in the agg...

Jim

[tejas9024]

sorry for the delay and Thanks for the responses.

We tried snapshot backups but we dint observe much difference in the backup durations.

Weekend full NDMP backups of about 35 TB usually take around 50 hours to complete. But during some instances they take more than 60 hours. We are unable to find the cause for this delay. we also checked if there were any scans run but there wasnt any. we are unsure on how to troubleshoot this issue. Please provide your inputs on this.