Here's the job file: [4ktest] fallocate=1 direct=1 rw=randwrite refill_buffers norandommap randrepeat=0 size=2G bs=4k iodepth=16 numjobs=16 runtime=60 group_reporting This job file was created based on a sample command line I was given: fio --filename=/oradata2/o18g2/IO/fiotest4 --direct=1 --rw=randwrite --refill_buffers --norandommap --randrepeat=0 --size=2G --bs=4k --iodepth=16 --numjobs=16 --runtime=60 --group_reporting --name=4ktest I think I understand enough that I need to run a rw=write job first before running the above job, correct? Mario Carino AIG Database Administrator IT Infrastructure | AIG Property Casualty 5 Wood Hollow Rd, 3rd Floor, Parsippany, NJ 07054 Tel +1 817 490-7145 | Cell +1 551 655-1443 | Direct +1 973 331-8558 (Mon/Wed/Thu) mario.carino@xxxxxxx | www.aig.com -----Original Message----- From: Jens Axboe [mailto:axboe@xxxxxxxxx] Sent: Thursday, April 10, 2014 11:34 AM To: Carino, Mario; fio@xxxxxxxxxxxxxxx Subject: Re: fio on AIX - fallocate error On 04/10/2014 09:30 AM, Carino, Mario wrote: > Hi Jens, > > Well, that's the thing. We can't get rid of the fallocate error > regardless of how we set fallocate= in the job file. We've tried > 0,1,none,posix and keep (which seems to be invalid now) and we ALWAYS > get the fallocate error. Sometimes the job continues to run, but the > output file is only created at zero byte length. Can you send me the job file you are using? -- Jens Axboe -- 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
