From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Subject: Re: Specify range and distribution of accesses References: <20160303161828.GC22164@kernel.dk> <20160303200446.GA2234@kernel.dk> From: Jens Axboe Message-ID: <56DDEDB4.6020704@kernel.dk> Date: Mon, 7 Mar 2016 14:08:04 -0700 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit To: Andrey Kuzmin , Jeff Furlong Cc: "fio@vger.kernel.org" List-ID: On 03/07/2016 02:02 PM, Andrey Kuzmin wrote: > On Mon, Mar 7, 2016 at 11:46 PM, Jeff Furlong wrote: >> Thanks for the suggestions and patches. Using the latest fio version, the JESD219 workload is possible: > > Nice. > >> >> # fio -version >> fio-2.6-27-gd283 >> >> # fio --name=JESD219 --ioengine=libaio --direct=1 --rw=randrw --norandommap --randrepeat=0 --rwmixread=40 --rwmixwrite=60 --iodepth=256 --size=100% --numjobs=4 --bssplit=512/4:1024/1:1536/1:2048/1:2560/1:3072/1:3584/1:4k/67:8k/10:16k/7:32k/3:64k/3 --random_distribution=zoned:50/5:30/15:20/80 --overwrite=1 --filename=/dev/nvme0n1 --group_reporting --runtime=5m --time_based --output=JESD219 >> >> A quick statistical analysis of the results shows: >> >> Found 20380582 IOs >> >> Found 39.9903152913% reads >> Found 60.0096847087% writes >> >> Found 4.00492979052% 512 >> Found 1.00495658073% 1024 >> Found 1.00079575745% 1536 >> Found 1.00046701316% 2048 >> Found 0.998764412125% 2560 >> Found 0.998043137335% 3072 >> Found 0.999520033334% 3584 >> Found 67.0145778958% 4096 >> Found 9.98662844859% 8192 >> Found 6.99898560306% 16384 >> Found 2.99961993235% 32768 >> Found 2.99271139558% 65536 >> >> Found 49.9895734086% 0-5% >> Found 30.0126463513% 5-20%% >> Found 19.99778024% 20-100% >> > > It hardly matters, but is still somewhat surprising to see that both > bs and zone split percentage are accurate only up to 5x10^-3. It tends to be more accurate with more IOs - for this case, it's 20 million, I guess you could assume that it'd be better. Generally it does get more accurate with more ios. But I'm mostly in the camp of "it hardly matters", it's close enough that you'd be hard pressed to complain about it. Fio does most of its math in integers, so we lose a bit of precision there, but it's a lot faster. That said, just a one-off in the calculations here could mean that it's an order of magnitude less accurate than it should. Maybe the above could be 10^-4 or 10^-5. It's so close that I'm finding it hard to locate the motivation to actually check and verify all that :-) -- Jens Axboe