Gluster users - Nov 2014 - [Gluster performance tuning]how to determine the value of performcance.cache-size

Hi  ,

How do I  determine  the value of  performance.cache-size option  suitable for
my glusterFS environment ? If I have large memory, can I set this value more
larger ?

Now I have 6 nodes which have 4 bricks.

Volume info is shown below:
Volume Name: gv0
Type: Distributed-Replicate
Volume ID: 8968985c-9272-4b03-8078-cf3ed91f9fc2
Status: Started
Number of Bricks: 12 x 2 = 24
Transport-type: tcp
Bricks:
Brick1: adf-db1:/export/disk1/brick
Brick2: adf-db2:/export/disk1/brick
Brick3: adf-db1:/export/disk2/brick
Brick4: adf-db2:/export/disk2/brick
Brick5: adf-db1:/export/disk3/brick
Brick6: adf-db2:/export/disk3/brick
Brick7: adf-db1:/export/disk4/brick
Brick8: adf-db2:/export/disk4/brick
Brick9: adf-rr1:/export/disk1/brick
Brick10: adf-rr2:/export/disk1/brick
Brick11: adf-rr1:/export/disk2/brick
Brick12: adf-rr2:/export/disk2/brick
Brick13: adf-rr1:/export/disk3/brick
Brick14: adf-rr2:/export/disk3/brick
Brick15: adf-rr1:/export/disk4/brick
Brick16: adf-rr2:/export/disk4/brick
Brick17: adf-fr1:/export/disk1/brick
Brick18: adf-fr2:/export/disk1/brick
Brick19: adf-fr1:/export/disk2/brick
Brick20: adf-fr2:/export/disk2/brick
Brick21: adf-fr1:/export/disk3/brick
Brick22: adf-fr2:/export/disk3/brick
Brick23: adf-fr1:/export/disk4/brick
Brick24: adf-fr2:/export/disk4/brick
Options Reconfigured:
network.ping-timeout: 20
performance.cache-size: 1024MB




BRs,
Jifeng.Li

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20141110/2d9cd1a7/attachment.html>

Hi Jifeng, as far as I know and from the tests I've run,
performance.cache-size is located on the client side (even though docs said
it could also be on the server side). Therefore be careful not to use more
memory than you have on your gluster clients.



2014-11-10 8:03 GMT-03:00 Jifeng Li <jifeng.li at ericsson.com>:
>  Hi  ,
>
>
>
> How do I  determine  the value of  performance.cache-size option
>  suitable for my glusterFS environment ? If I have large memory, can I set
> this value more larger ?
>
>
>
> Now I have 6 nodes which have 4 bricks.
>
>
>
> Volume info is shown below:
>
> Volume Name: gv0
>
> Type: Distributed-Replicate
>
> Volume ID: 8968985c-9272-4b03-8078-cf3ed91f9fc2
>
> Status: Started
>
> Number of Bricks: 12 x 2 = 24
>
> Transport-type: tcp
>
> Bricks:
>
> Brick1: adf-db1:/export/disk1/brick
>
> Brick2: adf-db2:/export/disk1/brick
>
> Brick3: adf-db1:/export/disk2/brick
>
> Brick4: adf-db2:/export/disk2/brick
>
> Brick5: adf-db1:/export/disk3/brick
>
> Brick6: adf-db2:/export/disk3/brick
>
> Brick7: adf-db1:/export/disk4/brick
>
> Brick8: adf-db2:/export/disk4/brick
>
> Brick9: adf-rr1:/export/disk1/brick
>
> Brick10: adf-rr2:/export/disk1/brick
>
> Brick11: adf-rr1:/export/disk2/brick
>
> Brick12: adf-rr2:/export/disk2/brick
>
> Brick13: adf-rr1:/export/disk3/brick
>
> Brick14: adf-rr2:/export/disk3/brick
>
> Brick15: adf-rr1:/export/disk4/brick
>
> Brick16: adf-rr2:/export/disk4/brick
>
> Brick17: adf-fr1:/export/disk1/brick
>
> Brick18: adf-fr2:/export/disk1/brick
>
> Brick19: adf-fr1:/export/disk2/brick
>
> Brick20: adf-fr2:/export/disk2/brick
>
> Brick21: adf-fr1:/export/disk3/brick
>
> Brick22: adf-fr2:/export/disk3/brick
>
> Brick23: adf-fr1:/export/disk4/brick
>
> Brick24: adf-fr2:/export/disk4/brick
>
> Options Reconfigured:
>
> network.ping-timeout: 20
>
> performance.cache-size: 1024MB
>
>
>
>
>
>
>
>
>
> BRs,
>
> Jifeng.Li
>
>
>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://supercolony.gluster.org/mailman/listinfo/gluster-users
>


-- 
Pavlik Salles Juan Jos?
Blog - http://viviendolared.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20141110/de7e7469/attachment.html>