[Gluster-devel] Best performance glusterfs v. 3.0.0 and samba

Roland Fischer roland.fischer at xidras.com
Tue Feb 16 09:27:27 UTC 2010

Hi Harshavardhana,

thank you for your reply. i will use the performance options for samba.

to your questions:

>  volume web-data-replicate
>     type cluster/replicate
>     subvolumes gfs-01-01 gfs-01-02
>  end-volume
>  volume readahead
>   type performance/read-ahead
>   option page-count 16              # cache per file  = (page-count x
>  page-size)
>   subvolumes web-data-replicate
>  end-volume
>>what is the client side and server side TOTAL ram ?. How many servers and
>>clients do you have?.  Coz having read-ahead count on 16 is no good for an
>>ethernet  link, you might be choking up the bandwidth unnecessarily.

we have 2 physikal xen server and 2 physikal gfs-server (one gfsserver with 48TB space - 24 x 2TB SATAII)
we use on xenserver 14 domUs
alle domUs are located on glusterfs. And now we try to share in domU1 another gfs-lun (gfs-partition) via samba

each of the physikal xen server have 2x Quadcore with 12GB ram
each of the physikal glusterfs server have 2x Quardcore (hyperthreading) with 12GB ram

xen - domU1 has 1 CPU and 2GB ram and shares another gfs-lun via samba

>>  Even with this we would need to know the backend disk performance with
>>  o-direct to properly analyse the cost of using buffering on server side to
>>  get better performance out of the system.

24 x 2TB - Western Digital WD RE4 - GP with 64MB Cache

Raid Controller = Areca - ARC-1280

Controller Name 	ARC-1280
Serial Number 	Y907CAAXAR800316
Main Processor 	800MHz IOP341 C1
CPU ICache Size 	32KBytes
CPU DCache Size 	32KBytes/Write Back
CPU SCache Size 	512KBytes/Write Back
System Memory 	256MB/533MHz/ECC

Thank you

More information about the Gluster-devel mailing list