[Lustre-discuss] How's this config for our Lustre setup?

[Peter Kjellström]

See my comments inline below.

On Friday, May 27, 2011 09:02:30 AM Kek wrote:
> Hi All
>  
> We're a research lab new to Lustre and are purchasing a small HPC cluster. 
> We wish to seek your comments and help on sizing the hardware. So far we
> plan to have the following:
> - 1 Head node, 2 x 6 core intel X5670, 32GB RAM, 2 x 300 GB SAS 10krpm HDDs
> - 1 MDS+MDT node, 2 x 6 core intel X5670, 24GB RAM, 2 x 500GB SATA 7.2krpm
> HDDs
> - 2 Storage nodes (OSS collocated with OST), each with: 2 x 4 core
> intel E5620, 24GB RAM, 14 x 600GB SAS 15krpm HDDs (raw 8.4TB)

I'd switch the SAS drives on the OSS-nodes for SATA and buy SAS on the MDT 
where IOPs matters more. I'd also consider a few more drives on the MDS like 
4-6 SAS in a raid10. Make sure your OSS-nodes use raid6.

> - 12 compute
> nodes, each with 2 x 6 core intel X5670.
> - Infiniband QDR connectivity
>  
> - Average file size: 20MB
> - Usage pattern: running parallel (MPI) models like WRF (weather
> model), POM (hydrodynamics model) etc 
> Questions:
> - Is the above hardware config all right?

Except for a general feeling of few and expensive components, yes. ;-)

> - What is the impact of putting the 2nd (failover) MDS+MDT on the Head

You filesystem will essentially be unavailalbe if one out of MDS + 2xOSS is 
down/dead so it doesn't make much HA sense to configure only the MDS with a 
fail-over (and it will require either drdb or shared external storage).

For a small config like this I'd say it's a no-brainer to keep it simple and 
stay away from fail-over stuff (we run a lustre with no failover and almost 
100 servers and even that has a reasonable availability).

> Node? - Is Lustre easy to maintain?
> - Is Lustre reasonably stable and problem free?

Compared to similar software I'd say so, yes. There is a quite nice admin-
guide that I recommend you read at:

 http://wiki.lustre.org/index.php/Lustre_Documentation

/Peter
  
> THanks for any tips or suggestions.
>  
> regards
> -chk
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20110527/8ead0b15/attachment.pgp>