Peter,
This is a really good description of how to do this. Would you mind adding
it to the FAQ? I'm sure there's an entry there for NUMBER_DATA_BUFFERS, etc.
http://www.backupcentral.com/cgi-bin/netbackup-fom
At 10:04 AM 5/8/2001 +1000, Marelas, Peter wrote:
>Tim,
>
>I tend to tune the buffers based on the maximum theoretical throughput
>of the tape drive in question.
>
>However to do so you need memory on the server.
>
>You've got it set to 16 and 64512 respectively.
>
>Therefore Netbackup will use 16*64512=1032192 bytes per tape drive
>in use.
>
>Assuming the maximum theoretical throughput of a drive is 6MB
>I would set SIZE_DATA_BUFFERS to the tape drives optimum block
>size (usually 64k, 128k or 256k).
>
>Lets assume its 64k (65536), divide 6291456 bytes (6MB) by 65536 bytes (64k)
>you
>get 96 NUMBER_DATA_BUFFERS. That should be the optimum setting.
>
>Regards
>Peter Marelas
>
>-----Original Message-----
>From: Tim Allen [mailto:tallen201 AT hotmail DOT com]
>Sent: Tuesday, 8 May 2001 1:06 AM
>To: Veritas-bu AT mailman.eng.auburn DOT edu
>Subject: [Veritas-bu] NBU Data Buffer sizes
>
>
>Hello.
>
>A database backup is taking a bit longer than expected, and I was
>considering bumping up the data buffers.
>SIZE_DATA_BUFFERS is currently 64512 and NUMBER_DATA_BUFFERS is 16.
>
>The bptm log has the following entries;
>
>04:12:47 [12400] <2> fill_buffer: [12399] socket is closed, waited for empty
>
>buf
>fer 462164 times, delayed 466388 times, read 40990432 Kbytes
>04:12:47 [12399] <2> write_data: waited for full buffer 1100 times, delayed
>1262
>2 times
>04:29:51 [12433] <2> fill_buffer: [12432] socket is closed, waited for empty
>
>buf
>fer 492546 times, delayed 498792 times, read 43451936 Kbytes
>04:29:51 [12432] <2> write_data: waited for full buffer 1206 times, delayed
>1131
>1 times
>06:52:49 [13024] <2> fill_buffer: [13023] socket is closed, waited for empty
>
>buf
>fer 426891 times, delayed 430477 times, read 37139232 Kbytes
>06:52:49 [13023] <2> write_data: waited for full buffer 591 times, delayed
>10367
>times
>07:45:18 [13096] <2> fill_buffer: [13095] socket is closed, waited for empty
>
>buf
>fer 543885 times, delayed 549783 times, read 47816800 Kbytes
>07:45:18 [13095] <2> write_data: waited for full buffer 930 times, delayed
>5408
>times
>08:13:28 [13810] <2> fill_buffer: [13808] socket is closed, waited for empty
>
>buf
>fer 38963 times, delayed 103462 times, read 2457696 Kbytes
>08:13:29 [13808] <2> write_data: waited for full buffer 0 times, delayed 0
>times
>08:48:43 [14012] <2> fill_buffer: [14011] socket is closed, waited for empty
>
>buf
>fer 38882 times, delayed 103166 times, read 2457696 Kbytes
>08:48:44 [14011] <2> write_data: waited for full buffer 0 times, delayed 0
>times
>09:23:17 [14150] <2> fill_buffer: [14149] socket is closed, waited for empty
>
>buf
>fer 37823 times, delayed 100564 times, read 2386784 Kbytes
>09:23:18 [14149] <2> write_data: waited for full buffer 0 times, delayed 0
>times
>09:54:18 [14302] <2> fill_buffer: [14301] socket is closed, waited for empty
>
>buf
>fer 34057 times, delayed 90351 times, read 2150432 Kbytes
>09:54:19 [14301] <2> write_data: waited for full buffer 3 times, delayed 32
>time
>s
>10:05:10 [13608] <2> fill_buffer: [13607] socket is closed, waited for empty
>
>buf
>fer 550854 times, delayed 557508 times, read 46860128 Kbytes
>10:05:10 [13607] <2> write_data: waited for full buffer 85 times, delayed
>2284 t
>imes
>10:18:30 [13885] <2> fill_buffer: [13884] socket is closed, waited for empty
>
>buf
>fer 421021 times, delayed 424891 times, read 39006624 Kbytes
>10:18:30 [13884] <2> write_data: waited for full buffer 597 times, delayed
>3158
>
>
>There are more delays waiting for an empty buffer than a full buffer so I
>was considering bumping the SIZE_DATA_BUFFERS to 129024.
>
>Hmmm... I guess the only reason for sending this e-mail was to ask if this
>sounds reasonable for what is in the bptm logs?
>
>I'll have to try it and find out if this improves anything next time.
>
>
>Tim Allen
>
>
>_________________________________________________________________
>Get your FREE download of MSN Explorer at http://explorer.msn.com
>
>_______________________________________________
>Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
>http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
>_______________________________________________
>Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
>http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
---
W. Curtis Preston
Principal Consultant for Storage Designs, your storage experts
Webmaster: http://www.backupcentral.com Phone: 760 631 7991
|
