[StarCluster] Starcluster creation issue with 100 + nodes

[Sumita Sinha]

Hello Team,

I have an account in amazon with instance limit of 2000.

I am facing issues in starting a 200 node cluster of micro instances (ebs
backed).

Its always kind of stuck at the step "*Waiting for SSH to come up on all
nodes..."*
*
*
After waiting for 1 hour 40 min i have taken a snapshot of the debug.log
file and attached with this mail.

Please let me know what is this step that takes so much time and cluster
keeps on waiting at this step.

The starting time of the 200 instances seems pretty fast.

Is there anything i need to take care of.

Any help would be appreciated.

On Wed, Dec 21, 2011 at 3:47 PM, Sumita Sinha <
sumita.sinha at claricetechnologies.com> wrote:

> Hello Justin,
>
> I am trying to create a cluster of 200 nodes.
>
> Have noticed the below step takes the maximum time and while creating
> cluster of  large number
> of nodes like 100 + , this is the step where it is always stuck however
> when i do a restart then it works.
>
> * Waiting for SSH to come up on all nodes...*
> *
> *
> Can you tell me what are the internal process this above step executes.
>
> I think during this step the ssh daemon starts running in all the nodes.
>
> Do we need to take care of something while creating cluster of large
> number of nodes.
> *
> *
>
> --
> Regards
> Sumita Sinha
>
>
>


-- 
Regards
Sumita Sinha
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/starcluster/attachments/20111222/6e313286/attachment-0001.htm
-------------- next part --------------
A non-text attachment was scrubbed...
Name: debug.log
Type: text/x-log
Size: 686083 bytes
Desc: not available
Url : http://mailman.mit.edu/pipermail/starcluster/attachments/20111222/6e313286/attachment-0001.bin