Unsolved
This post is more than 5 years old
16 Posts
0
932
Solaris 10 optimization settings for NetWorker 7.6.5
Hello,
We have a newly installed Oracle Sun Fire V445 server with 4 SPARC CPUs, 32 GB of RAM, 2 FC HBAs and 4 GbE ports in link aggregation (802.3ad) running Solaris 10 and we are wondering what are the optimization settings for maximum performance on this kind of machine running NetWorker 7.6.5.
This server controls 2 jukeboxes and a total of 15 drives (10 LTO5 and 5 LTO4) and we run NDMP (NetApp) via SAN as well as network backups (windows clients in majority) for a total of around 80 clients (Windows 2003, Windows 2008, Redhat Linux).
We backup around 40 TB of data each weekend and the goal is to reduce the backup window as much as possible in order to squeeze in cloning of this same 40 TB of data. In order of amounts, we have 30 TB of data which is backed up on NDMP via FC/SAN and 10 TB which is backed up over the GbE network. All clients have 1 GbE NICs but on the server we set up LACP link aggregation in order to have 4 Gb of total server bandwidth.
Please let me know what kernel, driver, network, or other Solaris 10 settings you can use in order to maximize backup throughput.
TimQuan
2 Intern
2 Intern
•
1.2K Posts
0
August 7th, 2013 18:00
NetWorker 7.6 Service Pack 2 Performance Optimization Planning Guide
Celio1
16 Posts
0
August 12th, 2013 04:00
Thanks for pointing out the official documentation, but I want to know what you guys use in the real world.
Another thing, on the Configuring TCP Networks and Network Firewalls for EMC NetWorker, P/N 300-005-739, REV A07, you find the following info regarding Solaris 10. But the A08 (latest) revision no longer contains that information. Does it mean it is no longer pertaining?
fx8282
38 Posts
0
August 12th, 2013 18:00
Are all the clients in the same network segment or are they separated by some kind of firewall, etc.
In case they are, then u need to run a test full backup run and observe any throughput/connectivity issues.
In case you do face any issues, please refer the above mentioned performance optimization guide or log a case with support to fine tune ur tcp keepalives, etc for a happier life