When backing up Unix servers, allowing multiple data streams will cause
Netbackup's stream discovery to create one job for each mounted filesystem.
However on Windows, it creates one job for each drive letter, regardless of the
number of mounted filesystems beneath each drive letter.
We are
We mount our volumes in a single location, so I found that adding a spec like
L:\volumes\vol* into the selection list had the desired effect: one job for
each mounted volume.
Thanks!
+--
|This was sent by [EMAIL PROTECTED]
Our numbers vary greatly according to the clients. We have very old Solaris
machines that backup at 50+ MB/sec. We have very new Windows machines where
throughput varies greatly between different volumes on the same machine (e.g. a
volume with 5 million tiny files backs up at 30 MB/sec via
PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf
Of sauderd
Sent: May 8, 2007 11:27 AM
To: VERITAS-BU@mailman.eng.auburn.edu
Subject: [Veritas-bu] Re: Tuning Solaris 10 for NetBackup Media Server
It was driver code to get the dladm command to work properly
with the Cisco switch
I spoke with the network guys in our shop and I got the following:
1. the IOS version is: 12.2, rev. 18 sfx 7 native
2. there is no minumum HW version
3. the hardware we're using is a Catalyst WS-X6748-GE-TX
4. they configured the ports as LCP in active mode
Being a Unix guys, I haven't the
Oh, I almost forgot to mention something. We got a specially created fix from
Sun to get the Solaris 10 part to work. A publicly available patch will
forthcoming soon.
___
Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu
It was driver code to get the dladm command to work properly with the Cisco
switch.
___
Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
The four onboard NICs as well as an extra 2-port NIC we installed are all GLDv3
devices. So, they take advantage of the new dladm command to do link
aggregation. We had the networking guys configure the 6 ports on the Cisco
switch (LACP protocol) and then we did the aggregation as follows:
Before we put our T2000's into service as Netbackup master servers, I
benchmarked their network throughput using ttcp and it was amazing. We pushed
5 simultaneous ttcp streams into the T2000 for a cumulative throughput of 495
MB/sec (that's megabytes). All of this was without it breaking a
