Re: Repro blocks/offsets in multiple runs

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Actually, I do see corruption in pattern file [ see line 2 ].

"/root/fbtest/fiotestfile write 4209246208 262144
/root/fiotestfile write 9908fio version 2 iolog
/root/fiotestfile add
/root/fiotestfile open
/root/fiotestfile write 1064566784 262144 "

So two questions:

1) How do I specify iolog file per job ?

2) How does the replay test file look like ? Is the one below fine
[based on original file in my previous mail] ?

[global]
description=Standard Performance test
direct=1
size=10g
filename=/root/fbtest/fiotestfile
numjobs=20
group_reporting

[randwrite_256]
stonewall
read_iolog=fio_pattern

Thanks,

Vikram

On Sat, Mar 10, 2012 at 9:20 AM, Vikram Seth <seth.vik@xxxxxxxxx> wrote:
> How do I specify a separate file for each job ?
>
> I have following simple test file [ see numjobs=20]. Will the logging
> work without corruption ?
>
> [global]
> description=Standard Performance test
> runtime=30
> direct=1
> size=10g
> filename=/root/fiotestfile
> numjobs=20
> group_reporting
>
> [randwrite_256]
> stonewall
> rw=randwrite
> bs=256k
> write_iolog=fio_pattern
>
> Vikram
>
> On Fri, Mar 9, 2012 at 11:29 PM, Daniel Buggie <danielabuggie@xxxxxxxxx> wrote:
>> I haven't tried them myself, but perhaps the write_iolog and
>> read_iolog options would work for you in this case?
>>
>>  write_iolog=str
>>              Write the issued I/O patterns to the specified file.
>> Specify a separate file for each job, otherwise the iologs will be
>> interspersed and the  file  may  be corrupt.
>>
>>       read_iolog=str
>>              Replay the I/O patterns contained in the specified file
>> generated by write_iolog, or may be a blktrace binary file.
>>
>> Daniel
>>
>> On Fri, Mar 9, 2012 at 2:23 PM, Vikram Seth <seth.vik@xxxxxxxxx> wrote:
>>> Hi,
>>>
>>> I am looking for an option that enables fio to use same set of
>>> blocks/offsets across multiple runs of the test.
>>> This will be useful if I want am trying to fix a perf issue and want
>>> to rerun the test to verify the patch. For such cases an exact repro
>>> of the i/o pattern is important.
>>>
>>> Thanks,
>>> Vikram.
>>> --
>>> 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
--
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


[Index of Archives]     [Linux Kernel]     [Linux SCSI]     [Linux IDE]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux SCSI]

  Powered by Linux