On Sun, Jul 17, 2011 at 01:51:11PM +0100, John Robinson wrote: > On 17/07/2011 13:22, Iustin Pop wrote: > >On Sun, Jul 17, 2011 at 01:11:19PM +0100, John Robinson wrote: > [...] > >># dd if=/dev/zero of=test bs=4096 count=262144 > >>262144+0 records in > >>262144+0 records out > >>1073741824 bytes (1.1 GB) copied, 2.5253 seconds, 425 MB/s > >> > >>And there's a badblocks running on another drive also on the ICH10. > >> > >>Having said that, I think mine's wrong too, I don't think my array > >>can really manage that much throughput. We should both be using more > >>realistic benchmarking tools like bonnie++: > > > >Or simply pass the correct flags to dd — like oflag=direct, which will > >make it do non-buffered writes. > > That's still not realistic: > > # dd if=/dev/zero of=test oflag=direct bs=4096 count=262144 > 262144+0 records in > 262144+0 records out > 1073741824 bytes (1.1 GB) copied, 117.434 seconds, 9.1 MB/s > > Because this time we're doing a read-modify-write for every 4K > block, or at least a write for every 4K block. Of course :) But at 4K speed, this is actually the speed of your array. > I can fix it up again > to work in stripe size amounts: > > # dd if=/dev/zero of=test oflag=direct bs=1572864 count=683 > 683+0 records in > 683+0 records out > 1074266112 bytes (1.1 GB) copied, 18.3198 seconds, 58.6 MB/s > > But it's still not realistic because real I/O does use buffers and Buffers is one thing; flushing to disk after a certain block size another, and that happens quite often, depending on workload. > doesn't work in magic sizes, so we should be using a more realistic > benchmarking tool like bonnie++. Honestly, I don't find bonnie++ a realistic tool. fio is a much better one; bonnie is quite old and inflexible. regards, iustin -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
