|
I personnaly do not use the merge-delay as it is
reported to be experimental. Further more. All my tests
demonstrated worth performances with merge-delay (I also tried
with 0.002 as you did). That said my setup is not perfect and I
don't want to misslead you. I have tried several times to get in
touch with Gabor Gombas who wrote ggaoed... but unfortanately all
my mails remained dead letters. If you're looking for performances maybe you can give vblade a try? (fork process or use bonding for multiple NICs) The first tests I made gave slightly better results with vblade than ggaoed concerning throuput: http://www.horoa.net/files/aoetest I think I used tiobench for thoose tests Le 07/07/2011 17:17, Lars Erik Dangvard Jensen a écrit : # Statistics for device aoestorage1-90-2-2000r6 read_cnt: 5235045 read_bytes: 21442744320 read_time: 9663.53 write_cnt: 0 write_bytes: 0 write_time: 0 other_cnt: 304 other_time: 0.0212562 io_slots: 1117752 io_runs: 96727 queue_length: 786443640 queue_stall: 1753 queue_over: 0 ata_err: 0 proto_err: 0I can get a very low "queue_stall" number by setting "merge-delay = 0.002" in ggaoed.conf, if not setting merge-delay queue_stall is extremely high.And "queue_over" is 0 when setting "queue-length = 1024". If I set queue-length to 64 I get extremely high queue_over number...# Statistics for device aoestorage1-90-2-2000r6 read_cnt: 16720474 read_bytes: 72916687360 read_time: 30601.7 write_cnt: 2101809 write_bytes: 8609009664 write_time: 8550.47 other_cnt: 1125131 other_time: 24.4878 io_slots: 4457695 io_runs: 397886 queue_length: 2795517659 queue_stall: 6826 queue_over: 0 ata_err: 46 proto_err: 0 I get some ata errors with an exported RAID6 array, but none for another RAID10 array on the same controller. Controller is: Adapter: 0 Product Name: Supermicro SMC2108 Memory: 512MB BBU: Present Serial No: What can cause ata_err ? (connected to several initiators all reporting ata error): aoe: ata error cmd=24h stat=41h from e90.2 end_request: I/O error, dev etherd/e90.2, sector 3898431360 Buffer I/O error on device etherd/e90.2, logical block 487303920 aoe: ata error cmd=24h stat=41h from e90.2 end_request: I/O error, dev etherd/e90.2, sector 3898431360 Buffer I/O error on device etherd/e90.2, logical block 487303920 ... I've checked the physical raid array on the target but this have no errors, and all disks are good. Can merge-delay or queue-length cause ata errors? ------------------------------------------------------------------------------ All of the data generated in your IT infrastructure is seriously valuable. Why? It contains a definitive record of application performance, security threats, fraudulent activity, and more. Splunk takes this data and makes sense of it. IT sense. And common sense. http://p.sf.net/sfu/splunk-d2d-c2 _______________________________________________ Aoetools-discuss mailing list Aoetools-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aoetools-discuss --
|
<<attachment: a_chapellon.vcf>>
------------------------------------------------------------------------------ All of the data generated in your IT infrastructure is seriously valuable. Why? It contains a definitive record of application performance, security threats, fraudulent activity, and more. Splunk takes this data and makes sense of it. IT sense. And common sense. http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________ Aoetools-discuss mailing list Aoetools-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aoetools-discuss
