I forgot to include the config params number_data_buffers 60 number_data_buffers_disk 60 size_data_buffers 524288 size_data_buffers 1048576
Bobby Williams 2205 Peterson Drive Chattanooga, Tennessee 37421 423-296-8200 _____ From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Bobby Williams Sent: Friday, November 30, 2007 7:16 AM To: 'Ed Wilts'; VERITAS-BU@mailman.eng.auburn.edu Subject: Re: [Veritas-bu] DSSU destaging performance? >From my bptm logs (using the term "copy 2" for destaging processes). I only do performance type queries on streams larger than 1GB. size speed 2,064,416 Kbytes at 156,859 Kbytes/sec 2,846,944 Kbytes at 141,879 Kbytes/sec 2,864,206 Kbytes at 130,381 Kbytes/sec 1,406,432 Kbytes at 125,496 Kbytes/sec 1,406,432 Kbytes at 116,736 Kbytes/sec 8,000,000 Kbytes at 110,228 Kbytes/sec 8,000,000 Kbytes at 99,475 Kbytes/sec 3,072,032 Kbytes at 89,119 Kbytes/sec 1,043,516 Kbytes at 81,340 Kbytes/sec System Model: IBM,9110-51A Machine Serial Number: 0685B1F Processor Type: PowerPC_POWER5 Number Of Processors: 4 (8 logical) Processor Clock Speed: 1648 MHz CPU Type: 64-bit Kernel Type: 64-bit LPAR Info: 1 06-85B1F Memory Size: 7936 MB AIX 5.3 NetBackup 5.0MP7 fiber attached LTO-3 IBM drives in an IBM library. All Fiber ports/sfp's are 4GB. Bobby Williams 2205 Peterson Drive Chattanooga, Tennessee 37421 423-296-8200 _____ From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Ed Wilts Sent: Thursday, November 29, 2007 4:18 PM To: VERITAS-BU@mailman.eng.auburn.edu Subject: [Veritas-bu] DSSU destaging performance? An LTO-3 drive needs to run at least 23MB/sec native and preferably over 80MB native (before compression). We're having difficulties consistently getting over 30 - and frequently get a lot lower - from our Solaris master server to our tape drives with NetBackup 6.0MP5. Our media servers aren't great either but they're going over ISLs although the ISLs aren't saturated (I'm trying to take them out of the picture for to eliminate the ISLs as a potential culprit). Are other people getting consistently better de-stage performance? I'm trying to determine if what we've got is unique to us or a more global issue. Symantec has been trying to help for over a year but we haven't been able to get our performance up to an acceptable level yet. If you're getting great performance, please drop me an email and let me know what your config is. Thanks, ./Ed -- Ed Wilts, RHCE, BCFP, BCSD Mounds View, MN, USA mailto:[EMAIL PROTECTED]
_______________________________________________ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu