NetBackup tuning question

[wbedour]

I am running NetBackup FP 4.5 on HP-UX 11.i, master server is on an HP L2000, and this media server is on an HP N4000 with a SureStore 20/700 library using LTO-1 drives. I created a policy that is doing a full backup, with cross mount points checked and limit jobs per policy to 4. Each mount point has about 175gb of data to be backed up. The policy kicks off 4 jobs as it should, but is taking 10 to 12 hours to complete.
I looked in my bptm logs and am seeing the following:
02:23:48.705 [24712] <2> write_data: waited for full buffer 2911 times, delayed 5980 times
02:24:17.000 [24714] <2> write_data: waited for full buffer 4019 times, delayed 4528 times

03:11:57.107 [26595] <2> write_data: waited for full buffer 1237 times, delayed 1487 times

04:03:31.939 [26594] <2> write_data: waited for full buffer 26659 times, delayed 72414 times

04:35:13.959 [4638] <2> write_data: waited for full buffer 33216 times, delayed 37370 times

04:57:24.506 [11931] <2> write_data: waited for full buffer 23206 times, delayed 25822 times

05:00:51.795 [26600] <2> write_data: waited for full buffer 114308 times, delayed 131449 times

05:06:11.186 [26608] <2> write_data: waited for full buffer 117920 times, delayed 138141 times

05:09:48.521 [25250] <2> fill_buffer: [25213] socket is closed, waited for empty buffer 186 times, delayed 188 times, read 52887552 bytes

05:09:48.535 [25213] <2> write_data: waited for full buffer 204 times, delayed 365 times

05:11:20.869 [25424] <2> fill_buffer: [25418] socket is closed, waited for empty buffer 1049 times, delayed 1143 times, read 300810240 bytes

05:11:20.882 [25418] <2> write_data: waited for full buffer 1161 times, delayed 1530 times

05:14:13.594 [1398] <2> fill_buffer: [1292] socket is closed, waited for empty buffer 0 times, delayed 0 times, read 144703488 bytes

05:14:13.612 [1292] <2> write_data: waited for full buffer 2782 times, delayed 3121 times

Do the above waits seem out of line? If so any suggestions on tuning this beast?

Thanks in advance

 

[PGPhantom]

The numbers look okay - I don't think you have anything to worry about there. Check out my FAQ - faq776-3124 for all buffer settings etc.

Are all your tape drives (All 20) connected to one media server? As a rule I never connect more than two LTO drives to any one media server - Depending on your architecture, Fibre etc? The NIC's - What make/model? What are your NIC settings and the Switch for the port s for all servers in question?

i.e. Nic's (If not using Fibre or Gigabit) should be manually set for 100/full at both nic and switch levels.

 

[wbedour]

We currently have only 8 tape drives in this library, all connected to the same media server. That is why I have the &quot;Limit Jobs per Policy&quot; parameter set at 4. The drives are scsi attached and I'm 99% sure we are running at 100/full.
We are currently not using multiplexing and don't plan on it if possible.
I checked for NET_BUFFER_SZ, NUMBER_DATA_BUFFERS & SIZE_DATA_BUFFERS and none exist so the NetBackup default is set for all these. Any suggestions on the initial setting of these to use as a base and then tune from there? Also, are these parameters dynamic or does NetBackup need to be cycled for the changes to take affect?
thanks

 

[PGPhantom]

Use the FAQ I provided - The settings there all work on HP UX 11.0 as we have a couple of media servers in the same environment. Part of the performance is your configuration - There is no possible way you can drive all the drives from one media server with a 100/full connection.

LTO is rated at 53GB/106GB per hour
100Mb/sec is 44GB/hour (IEEE 802.3u)
Even with only 4 drives running you could potentially write in excess of 200GB/hour - Your setup is not condusive to this. The HP recommendation is two LTO drives per server max and then only at a dual 100/mb nic or higher such as fibre or ether GB.

Seriously though - Reconsider your hardware layout.