Moving clusters

From Molecular Modeling Wiki

Revision as of 07:25, 28 August 2008

Warning

Due to the security problems, the SSH server keys will be regenerated on most cluster servers. This means that you can receive a warning similar to the following when trying to log in for the first time:

@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@    WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED!     @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
IT IS POSSIBLE THAT SOMEONE IS DOING SOMETHING NASTY!
Someone could be eavesdropping on you right now (man-in-the-middle attack)!
It is also possible that the RSA host key has just been changed.
The fingerprint for the RSA key sent by the remote host is
9a:a3:68:49:29:f6:a0:f4:c1:64:a0:fd:98:67:b2:67.
Please contact your system administrator.
Add correct host key in /root/.ssh/known_hosts to get rid of this message.
Offending key in /root/.ssh/known_hosts:53
RSA host key for iridium has changed and you have requested strict checking.
Host key verification failed.

To get rid of this warning and get access to the cluster, do the following:

1. Find a row that starts with Offending key ... in the text of the warning and remember line number after a colon (53 in the above example).
2. Open the ~/.ssh/known_hosts file delete the line which number was mentioned in the warning (in vi, press <5><3><shift-G><d><d> in case of the above example; replace numbers according to your needs).
3. Save the file and try to connect to the cluster.
4. If the connection fails again with the same or similar warning, repeat all steps.

Actual status

centrum

Up and running.

carol

Working on ...

marge

... to be moved.

teogate

... to be moved.

althea

Up and running.

lithium

Cluster is up and running, except for clients l22-l25 that need some repair.

helium

Cluster is up and running.

francium

Server is up and running.

There will be more maintenance work on server, so it can be restarted several times; meanwhile you can access your data.

The cluster is being checked.

The nodes will not be started before the infiniband switch is returned from a warranty repair - it can take two or three weeks.

iridium

Server is up and running.

There will be more maintenance work on server, so it can be restarted several times; meanwhile you can access your data.

The cluster is being checked.

Nodes will be started as soon as everything is done.

Need to connect new power cables from distant outlets, hope to finish this on Friday.

cobalt

Not running. Server will be turned on on Friday.

argon

Not running. Server will be turned on on Friday.

krypton

Not running. Server will be turned on on Friday.

radon

Not running. Server will be turned on on Friday.

palladium

Not running. Server will be turned on on Friday.

vanad

Discontinued.

titanium

Discontinued.

niob

Discontinued.