Hi Nick,
While you certainly can install the same 32 bit OS on heterogeneous CPUs
(including the native 64 bit ones) and put them under the same CE, this will
not allow the full use of features of new WNs. On the other hand, creation of
a whole new site just to accommodate new set of WNs may be an administrative
overkill.
IMHO, the preferred approach is to setup a new CE as a part of your site and
put new 64 bit WNs available through it. In the information system each CE is
described as a separate SubCluster, each having its own set of properties
fully describing the hardware and setup of WNs, which ideally suits the case
you describe.
Best regards, Antun
-----
Antun Balaz
Research Assistant
E-mail: [log in to unmask]
Web: http://scl.phy.bg.ac.yu/
Phone: +381 11 3713152
Fax: +381 11 3162190
Scientific Computing Laboratory
Institute of Physics, Belgrade, Serbia
-----
---------- Original Message -----------
From: Nick Vidiadakis <[log in to unmask]>
To: [log in to unmask]
Sent: Mon, 3 Sep 2007 09:14:36 +0300
Subject: [LCG-ROLLOUT] Different node pools and policy
> Good morning to all,
>
> What is the best approach to follow if there are two different node
> pools under the same site? If for example there are 10 AMD 64bit and
> 10 Xeon 32bit PCs, should someone install a new CE and declare a new
> site name, install a second CE in conjunction with the current one
> to manage only these WNs, use the current CE and create a proper
> policy using maui for example or use the same installation method
> for all WNs despite their differences in hardware? Is there any
> known implementation of a similar setup? If all nodes installed
> under the same CE use the same installation and O/S distribution
> (for example SL308 32bit), wouldn't it be harmful for the
> homogeneity of the specified cluster and for GRID in total?
>
> Nick Vidiadakis,
> GR-03-HEPNTUA
------- End of Original Message -------
|