Hi, When you report these issues could you cut the fio job file down to the smallest possible file that still reproduces the problem (i.e. removing any line in the job file should make the problem disappear)? On 3 May 2015 at 13:00, Srinivasa Chamarthy <chamarthy.raju@xxxxxxxxx> wrote: <snip> > # cat test1.fio > [global] > ioengine=libaio > size=1m > blocksize=1024k,1024k > iodepth=8 > iodepth_batch=4 > iodepth_low=2 > ;runtime=60 > ;time_based > filename=testfile_test1 > numjobs=1 > > [pattern_one_seq_write] > buffer_pattern=0xff > readwrite=write So the file size is 1024KBytes, the block size is 1024KBytes and you would like to queue 8 x 1024KByte (i.e. 8192KBytes) blocks in parallel at any one time? Is that correct and your intent? One immediate issue is that you aren't likely to achieve your I/O depth of 8 because OS restrictions on buffered I/O (see https://github.com/axboe/fio/blob/fio-2.2.7/HOWTO#L785 for warning about libaio and buffered I/O case). -- Sitsofe | http://sucs.org/~sits/ -- To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
