We are using an HNAS NFS mount as a basic disk storage unit for the purpose of catalog backups. when the full backup runs, the HNAS experiences significant packet loss per our NOC. The backup runs for about 6 hours and into the normal work day. Thus far I don't believe we've had any reported impact, but the packet loss has our NOC guys raising their eyebrows and worrying quite a bit.
i would like to find out if there is a way to limit the throughput to the NFS mount, from a Netbackup perspective. I'm checking with our BCE but thought i'd shoot up a flair here too in case someone has been down this road already.
i'm aware of being able to limit bandwidth for clients as outlined in that link you sent me... just not sure that is applicable for limiting bandwidth to a basic disk storage unit. also, it mentions that some backup types aren't affected... would NBU_Catalog be in the affected catagory? we would want to refrain from limiting bandwidth for the IP of our HNAS (because we are doing lots of NDMP backups, etc and don't want to slow those down) so i'm wondering if i could just create a unique hostname for the NFS mount to use, and apply the bandwidth limit to that only. thanks
Mike, unfortunately, NetBackup doesn't have any tools to limit bandwidth from media-server to storage unit because it's pretty uniq task. You can do it at OS level with some sort of traffic shapers.
@mikennedy - IP packet loss is a network issue, I my view you are looking in the wrong direction.
if you lower the bandwidth used for catalog backup, you are creating a workaround, not fixing the issues. You need to get a network technician to look at the issue.
we have max concurrent jobs set to 1 for the NFS mount storage unit
did you already try to do some tuning with
Maximum concurrent jobs storage unit setting ?
It could affect bandwidth usage positively.