[Linux-cluster] architecture discussion -- NFS clustering with iSCSI

Gordan Bobic gordan at bobich.net
Mon Jan 28 08:11:51 UTC 2008


Rainer Duffner wrote:
> 
> Am 28.01.2008 um 02:36 schrieb Terry:
> 
>>
>>
>> Good questions:
>> 1) Do you have concurrent writes to the same file from different nodes?
>>   1a) No
> 
> 
> Well, that's one of the things GFS is good at ;-)
> 
> 
>> 2) How many nodes do you have?
>>   2a) 3 to start, probably won't go beyond 12
>>
> OK, that's still in the range GFS can handle (AFAIK).

With an order of magnitude room for growth left.

>> I appreciate alternative ideas to NFS.  NFS could possibly introduce
>> performance issues (comments here appreciated).
> 
> 
> One problem might be that NFS was never supposed to run on  GBit-networks.
> Thus there is overhead.
> But, OTOH, the vendors I mentioned have managed to squeeze a lot of 
> performance out of NFS.
> It's also a question of optimizing/matching NFS clients and servers.

I've found that NFS v3 over UDP with large rsize,wsize and jumbo frames 
works pretty well.

>>   The majority of the
>> system is write.  I would say 80%.
>>
> 
> Do you have a lot of small files?
> Small files are usually what degrades GFS-performance.

I don't think small files are what kills it, it's lots of files that 
slow things down.

Gordan




More information about the Linux-cluster mailing list