The following piece of code (from iolog.c) suggests that fio should
only verify writes.
/*
* log a successful write, so we can unwind the log for verify
*/
void log_io_piece(struct thread_data *td, struct io_u *io_u)
{
. . .
ipo->numberio = io_u->numberio;
. . .
}
On Thu, Aug 29, 2013 at 5:19 PM, Juan Casse <jcasse@xxxxxxxxxx> wrote:
> I forgot to include the job file used in the test. Here it is.
>
> [global]
> readwrite=read
> randrepeat=1
> size=64k
> bs=8k
> ioengine=sync
> norandommap
> loops=2
> verify=meta
> verify_pattern=0xffffffffffffffff
> verify_dump=1
> continue_on_error=verify
>
> [jcasse_job]
>
>
> On Thu, Aug 29, 2013 at 5:09 PM, Juan Casse <jcasse@xxxxxxxxxx> wrote:
>> Jens,
>>
>> fio behaves in a way I was not expecting:
>>
>> Whenever the workload includes writes, fio will do the io and then
>> verify only the blocks that were written; this is good. However, when
>> the workload does only reads, the verify phase will verify the reads.
>>
>> So, the way the code is today, my changes to check numberio will work
>> only for workloads that do writes or read/writes, but not for
>> workloads that do only reads.
>>
>> I track the numberio in the pupulate_verify_io_u() function:
>> io_u->numberio = td->io_issues[io_u->ddir];
>> This function does not get called when doing reads.
>>
>> What are your thoughts?
>>
>>
>> On Thu, Aug 29, 2013 at 1:59 PM, Jens Axboe <axboe@xxxxxxxxx> wrote:
>>> On 08/29/2013 11:33 AM, Juan Casse wrote:
>>>> Currently, fio checks the block offset number in a block's header during
>>>> the verify phase.
>>>> We add a check for io number (numberio) to detect stale blocks.
>>>
>>> Any reason we can't do this unconditionally?
>>>
>>> --
>>> 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
