Re: Long checkpoint woe
In a long-checkpoint discussion, Neil Truby said the same slow dd results appeared on mirrored, unmirrored and internal disks of his Sun systems, raised a case with Sun, and asked Solaris SPARC users to run a 2k-block dd test. TBP argued the I/O rate was poor regardless of software. Roy Mercer posted his own dd read timings from a Solaris 9 Ultra-60. No cause was found.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
Are both primary and secondary raided the same way? are these the machines that are miles apart? The 3310 are directly attached (scsi?) aren't they?
"Nog" <noel@gfm.co.uk> wrote in message news:1143751527.646087.192390@e56g2000cwe.googlegroups.com... > Are both primary and secondary raided the same way? are these the > machines that are miles apart? The 3310 are directly attached (scsi?) > aren't they? The live one is exclusively RAID 1+0. The DR/DEV one has a mixture of RAID 1+0 and RAID 0. But the dd output is the same whether I use a mirrored logical volume, unmirrored LV, write directly to a device on the external array or write directly to one of the internal disks. I've raised a case with Sun. My hunch is they will come back and say it's expected behaviour, which is why i'd like anyone else with a Solaris SPARC box to so this: # time dd if=/dev/md/rdsk/d121 of=/dev/rdsk/c1t1d0s7 bs=2k 10240+0 records in 10240+0 records out real 1m1.85s user 0m0.03s sys 0m0.74s and let me know what they get, Thanks, Neil
Neil Truby wrote: > "Nog" <noel@gfm.co.uk> wrote in message > news:1143751527.646087.192390@e56g2000cwe.googlegroups.com... >> Are both primary and secondary raided the same way? are these the >> machines that are miles apart? The 3310 are directly attached (scsi?) >> aren't they? > > The live one is exclusively RAID 1+0. The DR/DEV one has a mixture of RAID > 1+0 and RAID 0. But the dd output is the same whether I use a mirrored > logical volume, unmirrored LV, write directly to a device on the external > array or write directly to one of the internal disks. > > I've raised a case with Sun. My hunch is they will come back and say it's > expected behaviour, What, crap I/O rates at 2k block size is "expected behaviour"? Humm, I wonder what Informix support could say "Humm, long checkpoints ... expected behaviour!" ... and which one would *you* accept? > which is why i'd like anyone else with a Solaris SPARC > box to so this: > > # time dd if=/dev/md/rdsk/d121 of=/dev/rdsk/c1t1d0s7 bs=2k > 10240+0 records in > 10240+0 records out > > real 1m1.85s > user 0m0.03s > sys 0m0.74s > > and let me know what they get, Thanks, Neil > > Does it matter? The I/O rate is crap - outside of any s/w apart from dd and an O/S kernel (and ... err ... the I/O sub-system) the I/O rate is crap.
time dd if=/dev/tempdbs01 of=/dev/null bs=2k count=10240 10240+0 records in 10240+0 records out real 0m2.77s user 0m0.05s sys 0m0.75s time dd if=/dev/tempdbs01 of=/tmp/del.out bs=2k count=10240 10240+0 records in 10240+0 records out real 0m3.11s user 0m0.08s sys 0m1.12s I don't have an extra raw device to write to but you should be able to compare these read times. uname -a SunOS 5.9 Generic_117171-17 sun4u sparc SUNW,Ultra-60