[Gluster-users] Question architecture of GlusterFS

[Vikas Gorur]

2008/10/22 Arend-Jan Wijtzes <ajwytzes at wise-guys.nl>:
> On Wed, Oct 22, 2008 at 08:49:27AM +0530, Basavanagowda Kanur wrote:
>> On Wed, Oct 22, 2008 at 3:49 AM, Mario Bonilla <hatrickmario at hotmail.com>wrote:
>>
>> > The architecture of GlusterFS configured as a RAID-1, using 2 nodes, the
>> > writing is done in a node and is replicated in the other node,
>> > question: the reading is done in parallel, since the two nodes where the
>> > information is replicated?
>>
>>
>> no.. reading is not done in parallel..
>> read is done from first available node.
>> if read from first node fails,
>>    then read is done through next available node.

> Is this behaviour required for correct operation? Does this mean that all
> clients are always reading from the same availble first node?
> Wouldn't the load be better distributed if you take a random or
> 'round robin' node to read from (which would seem trivial to implement)?

That is correct. Gowda gave you a simplified view of the operation.
The AFR translator supports a 'read-subvolume' option, which lets you
specify the node from where all the reads should be done. If you don't
specify the option, reads will be balanced in a nearly round-robin
fashion among all alive nodes.

Vikas Gorur
-- 
Engineer - Z Research
http://gluster.org/