I installed ruby as usual on the first node (s8270a17) in the cluster.
The question I have is "Can I just go to the next server in the
cluster and do "make install", or do I need to start back at the
./configure ?".
And along the same line, if I do start back at the ./configure will the
fact that it has already been run on a "different" server screw
anything up?
It depends on what your "identical linux servers" mean? If they are really identical you can omit the step "./configure" at the second server. Actually the default "./configure" makes prefix=/usr/local. I.e., You just put the source files on the shared file system and in fact you installed the ruby on local.
More importantly I think, no matter whether your servers are identical, it is right to issue the command "./configure" again on the shared file system from the second server. It will do nothing harmful, even if the 2nd server is a little different with the 1st node.
Shiwei
(The views expressed are my own and not necessarily those of Oracle and its affiliates.)
bradjpeek wrote:
ยทยทยท
I'm having second thoughts on the approach I've started down for
installing ruby on some Linux servers and would appreciate some
input...
I have two small clusters of identical Linux servers, each cluster
shares a file system named /scratch.
I have downloaded and extracted the ruby tar file to /scratch :
I installed ruby as usual on the first node (s8270a17) in the cluster.
The question I have is "Can I just go to the next server in the
cluster and do "make install", or do I need to start back at the
./configure ?".
And along the same line, if I do start back at the ./configure will the
fact that it has already been run on a "different" server screw
anything up?