From: Arun Krishnan (krishnan_at_ttck.keio.ac.jp)
Date: Wed Sep 12 2007 - 19:48:27 CDT
This happens most probably because your ssh public_key from the main (master
node) is not included in the "known_hosts" file under ~/.ssh . I would
suggest that you log in to all the nodes of your cluster from the master
node manually once... and then I suspect that the problem would go away.
Cheers,
Arun
On 9/11/07, Ahlam Al-Rawi <ahlamumali_at_gmail.com> wrote:
>
> Dear Namd user:
>
> We have no clue how the following error occurs. Can anybody help us:
>
> We Downloaded "standard" namd2 for Linux/amd64 and tried a few benchmark
> runs on a 8-way (16-core) Opteron-based node.
>
> We can run all the way up to 8 processes and the scaling looks quite
> amazing. However, when trying to run 16-process job using the
> following command, errors occurred (see below).
>
> [titan2:timing/1080-1nn] ~/NAMD_2.6_Linux-amd64/charmrun ++p 16 +
> +nodelist hosts ~/NAMD_2.6_Linux-amd64/namd2 equ11.conf
> ssh_exchange_identification: Connection closed by remote host
> ssh_exchange_identification: Connection closed by remote host
> ssh_exchange_identification: Connection closed by remote host
> ssh_exchange_identification: Connection closed by remote host
> ssh_exchange_identification: Connection closed by remote host
> ssh_exchange_identification: Connection closed by remote host
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
> Killed by signal 2.
>
> Thanks
> Ahlam
>
>
-- *********************************************** Arun Krishnan, Ph.D, Assistant Professor, Institute for Advanced Biosciences, Keio University, Center Building, Tsuruoka, Yamagata 997-0035 Japan Phone: +81 (0)235-29-0824 Email: krishnan_at_ttck.keio.ac.jp URL: http://www.iab.keio.ac.jp/~krishnan **********************************************
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:45:14 CST