whirr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alex Heneveld <Alex.Henev...@CloudsoftCorp.com>
Subject Re: Dying because - net.schmizz.sshj.transport.TransportException: Broken transport; encountered EOF
Date Wed, 26 Oct 2011 12:28:13 GMT


I've seen the same errors you are seeing with that distro on Amazon and 
they are benign.  These ubuntu images have a race creating ssh server 
keys, so sometimes at startup sshd will abort connections which 
jclouds/sshj moans about, but it rights itself within a few seconds.

The machines listed at [1] seem to be the ones jclouds currently 
returns, recent official Ubuntu contributed by Canonical.  It looks like 
this is what you're hard-coding so I'd stick with what you're already 
doing and disregard error messages from sshj that occur soon after 
machine startup.

(That said, we're not entirely certain the images are okay; we have 
tentative reports of other ssh problems with the 32-bit machines, but 
haven't been able to isolate them.  If there are problems they are 
highly intermittent -- but if you have any other issues please let us know.)


[1]  http://uec-images.ubuntu.com/query/lucid/server/released.current.txt

On 26/10/2011 11:06, Andrei Savu wrote:
> On Wed, Oct 26, 2011 at 1:01 PM, Paolo Castagna 
> <castagna.lists@googlemail.com <mailto:castagna.lists@googlemail.com>> 
> wrote:
>     Should I remove whirr.image-id=eu-west-1/ami-ee0e3c9a from my recipe?
> IMHO I think it's better to specify an AMI so that you can avoid 
> surprises. We are using this approach for integration testing so that 
> we don't have to customise the .properties file when switching from 
> aws-ec2 to cloudservers.

View raw message