[Gluster-users] read and write speeds

[Gerald Brandt]

Just to follow up, my speeds read and write speeds have not balanced out yet.

These are the settings on my NFS gluster system:

# gluster volume info

Volume Name: datastore
Type: Replicate
Status: Started
Number of Bricks: 1
Transport-type: tcp
Bricks:
Brick1: 10.1.40.2:/nfs/disk3
Options Reconfigured:
performance.read-ahead: off
performance.io-cache: on
performance.write-behind-window-size: 4MB
performance.io-thread-count: 64
performance.cache-size: 134217728
nfs.trusted-sync: on
nfs.port: 2049


----- Original Message -----
> Hi,
> 
> I'm testing gluster/nfs for replacement of an existing DRBD/iSCSI
> system.
> 
> Speed tests show gluster NFS to be pretty close to iSCSI, but I have
> some questions.
> 
> If I do a sequential write of data, I get ~118 MB/s.  A sequential
> read of data gets about 65 MB/s.
> 
> If I do a sequential read and write at the same time, write speed
> drops to ~100 MB/s while read speed drops to about 10 MB/s.
> 
> Does anyone know why this occurs, and is there anything I can do to
> speed up the read speeds, maybe make reads and writes a bit more
> balanced?
> 
> Gerald
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
>