[Gluster-devel] Last git tree memory problem ?

Shehjar Tikoo shehjart at zresearch.com
Fri Apr 10 12:52:35 UTC 2009


Hi

We've just checked-in 2 commits that reduce per-thread memory usage
as well as the default number of threads. The first commit should result
in lower memory usage in your vol file with 64 threads. Please let us
know the memory usage for the latest git with the same server side config.

You can see a brief description of the two commits here:
http://git.savannah.gnu.org/cgit/gluster.git/log/

Thanks
Shehjar

nicolas prochazka wrote:
> hello again, after more test, glusterfd process memory grows with 
> ption thread-count 64 , perhaps is it normal ?
> 
> ( count = 4  => 300 Mo , 16 => 600 Mo , 64 => 2.5Go )
> 
> Regards, Nicolas Prochazka
> 
> 
> 2009/4/10 nicolas prochazka <prochazka.nicolas at gmail.com>:
>> Hello, server and client conf file Regards, Nicolas Prochazka.
>> 
>> 
>> volume brickless type storage/posix option directory 
>> /mnt/disks/export end-volume
>> 
>> volume brickthread type features/posix-locks option mandatory-locks
>>  on          # enables mandatory locking on all files subvolumes 
>> brickless end-volume
>> 
>> volume brick type performance/io-threads option thread-count 64 
>> subvolumes brickthread end-volume
>> 
>> 
>> volume server type protocol/server subvolumes brick option 
>> client-volume-filename /etc/glusterfs/Gglusterfs-client.vol option 
>> transport-type tcp option auth.addr.brick.allow 10.98.98.* 
>> end-volume
>> 
>> 
>> 
>> 
>> --------------------------------------------------------------------
>>  volume brick_10.98.98.1 type protocol/client option transport-type
>>  tcp/client option transport-timeout 120 option remote-host 
>> 10.98.98.1 option remote-subvolume brick end-volume
>> 
>> 
>> volume brick_10.98.98.2 type protocol/client option transport-type 
>> tcp/client option transport-timeout 120 option remote-host 
>> 10.98.98.2 option remote-subvolume brick end-volume
>> 
>> 
>> volume last type cluster/replicate subvolumes brick_10.98.98.2 
>> brick_10.98.98.1 option favorite-child brick_10.98.98.1 end-volume
>> 
>> 
>> volume iothreads type performance/io-threads option thread-count 64
>>  subvolumes last end-volume
>> 
>> volume io-cache type performance/io-cache option cache-size 2048MB 
>> # default is 32MB option page-size  128KB             #128KB is 
>> default option option cache-timeout 2  # default is 1 subvolumes 
>> iothreads end-volume
>> 
>> volume writebehind type performance/write-behind option block-size 
>> 128KB # default is 0bytes option cache-size 512KB option 
>> flush-behind off      # default is 'off' subvolumes io-cache 
>> end-volume
>> 
>> 
>> 2009/4/10 Shehjar Tikoo <shehjart at zresearch.com>:
>>> I am looking into this. In the meantime, It'd help very much if 
>>> you can post the vol file.
>>> 
>>> Thanks Shehjar
>>> 
>>> nicolas prochazka wrote:
>>>> hello, I just try last git tree, and it seems that's server 
>>>> process ( glusterfsd ) takes a lot of memory ( >2 Go) whereas 
>>>> with same config file and other version, glusterfsd process 
>>>> take 300Mo
>>>> 
>>>> J've juste change , option thread-count 4 to option 
>>>> thread-count 64 ( it seems that now thread count must be 
>>>> between 32-512 ?)
>>>> 
>>>> Regards, Nicolas Prochazka
>>>> 
>>>> 
>>>> _______________________________________________ Gluster-devel 
>>>> mailing list Gluster-devel at nongnu.org 
>>>> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>>>> 
>>> 
> 






More information about the Gluster-devel mailing list