[zfs-discuss] zpool iostat and iostat discrepancy
Hi, Does anyone know the difference between zpool iostat and iostat? dd if=/dev/zero of=/test/test1/trash count=1 bs=1024k;sync pool only shows 236K IO and 13 write ops. whereas iostat shows a correctly meg of activity. zpool iostat -v test 5 capacity operationsbandwidth pool used avail read write read write -- - - - - - - test1.14M 100G 0 13 0 236K c8t60060E800475F50075F50525d0 182K 25.0G 0 4 0 36.8K c8t60060E800475F50075F50526d0 428K 25.0G 0 4 0 87.7K c8t60060E800475F50075F50540d0 558K 50.0G 0 4 0 111K -- - - - - - - iostat -xnz [devices] 5 extended device statistics r/sw/s kr/s kw/s wait actv wsvc_t asvc_t %w %b device 2.46.06.8 88.2 0.0 0.00.01.0 0 0 c8t60060E800475F50075F50540d0 2.45.46.8 37.0 0.0 0.00.00.9 0 0 c8t60060E800475F50075F50526d0 2.45.06.8 112.0 0.0 0.00.00.9 0 0 c8t60060E800475F50075F50525d0 dtrace also concurs with iostat device bytes IOPS == /devices/scsi_vhci/s...@g60060e800475f50075f50525:a 224416 35 /devices/scsi_vhci/s...@g60060e800475f50075f50526:a 486560 37 /devices/scsi_vhci/s...@g60060e800475f50075f50540:a 608416 33 Thanks -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool iostat and iostat discrepancy
On 06/20/09 11:14, tester wrote: Hi, Does anyone know the difference between zpool iostat and iostat? dd if=/dev/zero of=/test/test1/trash count=1 bs=1024k;sync pool only shows 236K IO and 13 write ops. whereas iostat shows a correctly meg of activity. The zfs numbers are per second as well. So 236K * 5 = 1180K zpool iostat -v test 1 would make this clearer. The iostat output below also shows 237K (88+37+112) being written per second. I'm not sure why any reads occurred though. When I did a quick experiment there were no reads. Enabling compression gives much better numbers when writing zeros! Neil. zpool iostat -v test 5 capacity operationsbandwidth pool used avail read write read write -- - - - - - - test1.14M 100G 0 13 0 236K c8t60060E800475F50075F50525d0 182K 25.0G 0 4 0 36.8K c8t60060E800475F50075F50526d0 428K 25.0G 0 4 0 87.7K c8t60060E800475F50075F50540d0 558K 50.0G 0 4 0 111K -- - - - - - - iostat -xnz [devices] 5 extended device statistics r/sw/s kr/s kw/s wait actv wsvc_t asvc_t %w %b device 2.46.06.8 88.2 0.0 0.00.01.0 0 0 c8t60060E800475F50075F50540d0 2.45.46.8 37.0 0.0 0.00.00.9 0 0 c8t60060E800475F50075F50526d0 2.45.06.8 112.0 0.0 0.00.00.9 0 0 c8t60060E800475F50075F50525d0 dtrace also concurs with iostat device bytes IOPS == /devices/scsi_vhci/s...@g60060e800475f50075f50525:a 224416 35 /devices/scsi_vhci/s...@g60060e800475f50075f50526:a 486560 37 /devices/scsi_vhci/s...@g60060e800475f50075f50540:a 608416 33 Thanks ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zpool iostat and iostat discrepancy
Neil, Thanks. That makes sense. May be man page for zpool can say that it is a rate as iostat man page does. I think reads are from the zpool iostat command itself. zpool iostat doesn't capture that. Thanks -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
