[Gluster-users] Speed up heal performance

[Vijay Bellur]

On Tuesday 13 October 2015 09:03 AM, Atin Mukherjee wrote:
>
>
> On 10/13/2015 08:13 AM, Lindsay Mathieson wrote:
>>
>> On 13 October 2015 at 11:56, Atin Mukherjee <atin.mukherjee83 at gmail.com
>> <mailto:atin.mukherjee83 at gmail.com>> wrote:
>>
>>      Good news is we already have a WIP patch review.glusterd.org/10851
>>      <http://review.glusterd.org/10851> to introduce multi threaded shd.
>>
>>
>> Thanks Atin, no need for the RFE then?
> https://bugzilla.redhat.com/show_bug.cgi?id=1221737 - as I mentioned
> earlier it was already opened.
>>
>> I notice the patch mainly discusses improvingthe performanceof healing
>> many small files, will it help with large VM files as well?
> AFAIK, it would help large files as well, but you would see a
> significant improvement specially for small files IMO. AFR team can
> correct me otherwise.


[1] has details about the improvements being planned for both data and 
entry self-heals. While data self-healing improvements mostly help large 
files, granular entry self-heal can help when you have a large number of 
files in a single directory (typically seen with small files).

multi-threaded self-heal daemon should help both small & large files by 
parallelizing the number of heals.

Regards,
Vijay


[1] 
http://review.gluster.org/#/c/12257/1/in_progress/afr-self-heal-improvements.md