fio usage question

fio usage question

[John OCallaghan]

Hi All

A question on usage of fio. I would like to be able to spin up more than
one fio instance running on the same platform concurrently.

For example, I would like to run one fio instance in time_based mode
doing some stuff. And after a fixed period of time then start a non-time
based fio instance doing some other stuff.

I have tried it out and to my untrained eyes it looks like its going OK.
But I'd love to hear from others with more experience in this area. Is
this something that is recommended?

Will my data skew in some way because fio expects there to be only one
instance of itself playing at any one time?

Any advice would be much appreciated!

John

-----

The information contained in this e-mail and in any attachments is confidential and is designated solely for the attention of the intended recipient(s). If you are not an intended recipient, you must not use, disclose, copy, distribute or retain this e-mail or any part thereof. If you have received this e-mail in error, please notify the sender by return e-mail and delete all copies of this e-mail from your computer system(s). Please direct any additional queries to: communications@s3group.com. Thank You. Silicon and Software Systems Limited (S3 Group). Registered in Ireland no. 378073. Registered Office: South County Business Park, Leopardstown, Dublin 18.

Re: fio usage question

[David Nellans]

On 08/18/2014 10:03 AM, John OCallaghan wrote:
>
> Hi All
>
> A question on usage of fio. I would like to be able to spin up more than
> one fio instance running on the same platform concurrently.
>
> For example, I would like to run one fio instance in time_based mode
> doing some stuff. And after a fixed period of time then start a non-time
> based fio instance doing some other stuff.
>
> I have tried it out and to my untrained eyes it looks like its going OK.
> But I'd love to hear from others with more experience in this area. Is
> this something that is recommended?
>
> Will my data skew in some way because fio expects there to be only one
> instance of itself playing at any one time?
>
> Any advice would be much appreciated!
>
Different processes shouldn't run into any issues unless you're doing a 
job with
verification in which case you'll end up with verify errors due to jobs 
overwriting each
other's data.

Unless you want to use multiple processes for some reason versus jobs 
inside a single
process - you can delay one of your multiple jobs starting using the 
startdelay=N
parameter in the jobfile.

Re: fio usage question

[John OCallaghan]

Thanks David, am not using verify but am curious how this can get affected.

Would it be that job configs that use the same file names would clobber
one anothers data on disk? I could see how that would be an issue for
multi process fio instances alright.




On 18/08/14 16:38, David Nellans wrote:
> On 08/18/2014 10:03 AM, John OCallaghan wrote:
>>
>> Hi All
>>
>> A question on usage of fio. I would like to be able to spin up more than
>> one fio instance running on the same platform concurrently.
>>
>> For example, I would like to run one fio instance in time_based mode
>> doing some stuff. And after a fixed period of time then start a non-time
>> based fio instance doing some other stuff.
>>
>> I have tried it out and to my untrained eyes it looks like its going OK.
>> But I'd love to hear from others with more experience in this area. Is
>> this something that is recommended?
>>
>> Will my data skew in some way because fio expects there to be only one
>> instance of itself playing at any one time?
>>
>> Any advice would be much appreciated!
>>
> Different processes shouldn't run into any issues unless you're doing a
> job with
> verification in which case you'll end up with verify errors due to jobs
> overwriting each
> other's data.
> 
> Unless you want to use multiple processes for some reason versus jobs
> inside a single
> process - you can delay one of your multiple jobs starting using the
> startdelay=N
> parameter in the jobfile.
-----

The information contained in this e-mail and in any attachments is confidential and is designated solely for the attention of the intended recipient(s). If you are not an intended recipient, you must not use, disclose, copy, distribute or retain this e-mail or any part thereof. If you have received this e-mail in error, please notify the sender by return e-mail and delete all copies of this e-mail from your computer system(s). Please direct any additional queries to: communications@s3group.com. Thank You. Silicon and Software Systems Limited (S3 Group). Registered in Ireland no. 378073. Registered Office: South County Business Park, Leopardstown, Dublin 18.

Re: fio usage question

[David Nellans]

you just need to adequately keep the sectors being written by each job 
segregated.
filenames, blkdev ranges, read only jobs - a variety of ways you could 
keep them from
trouncing on each other depending what your job is doing.

if you don't then you'll have a race where one process can change the 
data on disk
before the other job has verified,  leading to false verify failures.

On 08/18/2014 10:46 AM, John OCallaghan wrote:
>
> Thanks David, am not using verify but am curious how this can get affected.
>
> Would it be that job configs that use the same file names would clobber
> one anothers data on disk? I could see how that would be an issue for
> multi process fio instances alright.
>
>
>
>
> On 18/08/14 16:38, David Nellans wrote:
>> On 08/18/2014 10:03 AM, John OCallaghan wrote:
>>> Hi All
>>>
>>> A question on usage of fio. I would like to be able to spin up more than
>>> one fio instance running on the same platform concurrently.
>>>
>>> For example, I would like to run one fio instance in time_based mode
>>> doing some stuff. And after a fixed period of time then start a non-time
>>> based fio instance doing some other stuff.
>>>
>>> I have tried it out and to my untrained eyes it looks like its going OK.
>>> But I'd love to hear from others with more experience in this area. Is
>>> this something that is recommended?
>>>
>>> Will my data skew in some way because fio expects there to be only one
>>> instance of itself playing at any one time?
>>>
>>> Any advice would be much appreciated!
>>>
>> Different processes shouldn't run into any issues unless you're doing a
>> job with
>> verification in which case you'll end up with verify errors due to jobs
>> overwriting each
>> other's data.
>>
>> Unless you want to use multiple processes for some reason versus jobs
>> inside a single
>> process - you can delay one of your multiple jobs starting using the
>> startdelay=N
>> parameter in the jobfile.
> -----
>
> The information contained in this e-mail and in any attachments is confidential and is designated solely for the attention of the intended recipient(s). If you are not an intended recipient, you must not use, disclose, copy, distribute or retain this e-mail or any part thereof. If you have received this e-mail in error, please notify the sender by return e-mail and delete all copies of this e-mail from your computer system(s). Please direct any additional queries to: communications@s3group.com. Thank You. Silicon and Software Systems Limited (S3 Group). Registered in Ireland no. 378073. Registered Office: South County Business Park, Leopardstown, Dublin 18.
>
> --
> To unsubscribe from this list: send the line "unsubscribe fio" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html