Configure ssh for MPI - Revision history

Jprenaud at 11:01, 21 August 2007

2007-08-21T11:01:27Z

2007-08-21T10:56:45Z

2007-08-21T10:51:21Z

ssh-agent

2007-08-21T10:50:57Z

Copying the authorised key on each host

2007-08-21T10:50:27Z

Copying the authorised key on each host

2007-08-21T10:50:03Z

Adding a bit of security

2007-08-21T10:49:48Z

Adding a bit of security

2007-08-21T10:49:04Z

Specification of the key to use

2007-08-21T10:48:39Z

Copy the public key to the slaves

2007-08-21T10:48:10Z

Shared home directory between the master and the slaves

← Older revision		Revision as of 11:01, 21 August 2007
Line 1:		Line 1:
	To be able to run a parallel computation on a network of computers via MPI, one has to be able to log in to any of the machines without having to enter a password. This can be achieved easily using secure shell key authentication. This article describes the methodology to setup quickly ssh key authentication for MPI applications.		To be able to run a parallel computation on a network of computers via MPI, one has to be able to log in to any of the machines without having to enter a password. This can be achieved easily using secure shell key authentication. This article describes the methodology to setup quickly ssh key authentication for MPI applications.
		+
		+	This is useful even if you don't run MPI jobs actually but you often access machines via secure shell.

	= Principle of key authentication=		= Principle of key authentication=

@@ Line 99: / Line 99: @@
 = ssh-agent =
-<code>ssh-agent</code> is a program which keeps pass-phrases in memory and re-uses them automatically when required by the secure shell. <code>ssh-agent</code> just outputs some variables and to start it, you need to "evaluate" it:
+<code>ssh-agent</code> is a program which keeps pass-phrases in memory and re-uses them automatically when required by the secure shell. <code>ssh-agent</code> just outputs some variables in shell synatx and to start it, you need to "evaluate" it:
 <code><pre>

@@ Line 96: / Line 96: @@
 * If you are using a blank pass-phrase and you can log into each slave without user input, that's it! You don't need to do anything else regarding ssh.
-* if you are not using a blank pass-phrase, you will need to use a pass-phrase caching program called <code>ssh-agent</code>, see below.
+* If you are not using a blank pass-phrase, you will need to use a pass-phrase caching program called <code>ssh-agent</code>, see below.
 = ssh-agent =

@@ Line 89: / Line 89: @@
 = Copying the authorised key on each host =
-Now that slave1 is fully configured, you can copy the configuration to each host. Simply copy the <code>authorized_keys</code> file in <code.ssh/</code> on slave1 to the same location on master, slave2, slave3, etc... Note that if your home directory is shared between master/slaves, you don't have anything to do here.
+Now that slave1 is fully configured, you can copy the configuration to each host. Simply copy the <code>authorized_keys</code> file in <code>.ssh/</code> on slave1 to the same location on master, slave2, slave3, etc... Note that if your home directory is shared between master/slaves, you don't have anything to do here.
 '''To check: do we really need the authorized_keys entry on master for MPI purposes??'''

@@ Line 79: / Line 79: @@
 = Adding a bit of security =
-If you know that you will only be using master to start your MPI jobs, you can tighten a bit the configuration by requestion that the key <code>id_rsa.mpi.pub</code> is only an authorised key when logging in from master. This is done by editing the <code>authorised_keys</code> on each slave as shown below:
+If you know that you will only be using master to start your MPI jobs, you can tighten a bit the configuration by requestion that the key <code>id_rsa.mpi.pub</code> is only an authorised key when logging in from master. This is done by editing the <code>authorised_keys</code> on the slave as shown below:
 <code><pre>

← Older revision	Revision as of 10:56, 21 August 2007
(No difference)

@@ Line 66: / Line 66: @@
 As we have renamed the key, we need to tell the ssh program that we intent to use this non-default key. This is done by creating a file named <code>config</code> in the <code>.ssh</code> folder on master which contains the following lines:
 <code><pre>
 IdentityFile ~/.ssh/id_rsa.mpi
 </pre></code>

@@ Line 37: / Line 37: @@
 </pre></code>
-= Copy the public key to the slaves =
+= Copy the public key to a slave =
 The public key now needs to be copied into a file called <code>authorized_keys</code> in the <code>.ssh/</code> folder on each slave. The methods is slightly different if your home directory is shared between the master/slaves or not.

@@ Line 62: / Line 62: @@
 master $ cat .ssh/id_rsa.mpi.pub >> .ssh/authorized_keys
 </pre></code>
 = Specification of the key to use =