11-14-2011 07:51 AM
Does anyone use client throttling? We have a handful of clients which are having issues dropping connection which the Admin's say is caused because of netbackup. We've upgraded our cisco switches so they believe more bandwidth is getting used by the backup jobs. I don't agree but they have it in there head that we should throttle the backup on certain clients.
I've tried to throttle the backup on one client. For example under Master server properties....Bandwidth....I've configure (the client in question)
Bandwidth throttling setting for the following range of hosts:
From Host 10.160.3.222 To Host: 10.160.3.222 Bandwidth 7000
From my understanding, this is supposed to only throttle the bandwidth for this client. What happens is it causes all backups to run slowly. I've also moved the client in question to a single stream backup.
Am I setting this up correctly?
Solved! Go to Solution.
11-14-2011 09:43 AM
Bug in NBU 7.1:
Extract from Release Notes :
■ Do not use a LIMIT_BANDWIDTH (IPv4) configuration or a
THROTTLE_BANDWIDTH (IPv6) configuration in the initial release of
NetBackup 7.1. If you use either of these configuration options, you can
encounter a sever performance degradation.
An EEB will be available to fix the issues.
As far as Tuning on NBU clients and servers are concerned, have a look at this excellent article:
https://www-secure.symantec.com/connect/articles/tuning-windows-2003-and-2008-symantec-netbackup
11-14-2011 09:43 AM
Bug in NBU 7.1:
Extract from Release Notes :
■ Do not use a LIMIT_BANDWIDTH (IPv4) configuration or a
THROTTLE_BANDWIDTH (IPv6) configuration in the initial release of
NetBackup 7.1. If you use either of these configuration options, you can
encounter a sever performance degradation.
An EEB will be available to fix the issues.
As far as Tuning on NBU clients and servers are concerned, have a look at this excellent article:
https://www-secure.symantec.com/connect/articles/tuning-windows-2003-and-2008-symantec-netbackup
11-14-2011 09:49 AM
The issue Marianne points out was fixed in 7.1.0.1:
http://www.symantec.com/docs/DOC3941
Etrack Incident: 2314618
■ Contained in: NBLU_7.1.0.1.UNIX NBLU_7.1.0.1.WIN NB_7.1.0.1
NB_7.1.0.1.winnt.IA64 NB_7.1.0.1.winnt.x64 NB_7.1.0.1.winnt.x86
NB_CLT_7.1.0.1
■ Description:
This release contains fixes to address the following issues:
■ A memory leak with vnet_fqdn() (refer also to ET2291609).
■ Problems that were identified with LIMIT_BANDWIDTH (refer also to
ET2273461).
■ Memory leak fixes for the host cache (refer also to ET2294837).
11-14-2011 11:13 AM
Thank you!!! I'm getting ready to apply 7.1.0.2 hopefully this week, so I can do some further testing once that's applied. I will pass that article on to our Windows admins. We are also having issues on Solaris clients but they said changing them to single data streams has improved performance.