SCIENTIFIC-LINUX-USERS Archives

December 2005

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Kuo Kan Liang <[log in to unmask]>
Reply To:
Kuo Kan Liang <[log in to unmask]>
Date:
Wed, 7 Dec 2005 00:48:16 +0800
Content-Type:
text/plain
Parts/Attachments:
text/plain (47 lines)
Hi, all,

The problem is kind of long and complicated.

I installed SL4.1 i386 on my cluster. All of the computers come with 2 xeon.
The server is also connected to a SATA-SCSI raid through an LSI scsi card.
At that time I could not connect to the raid successfully and I thought 
it was the
problem of the OS, so, according to some suggestions, I changed the OS to
SL 4.1 x86_64 version. But it turns out that the disk problem was mainly 
due to
old SCSI card firmware. After updating the firmware, the raid problem was
solved.

However, when I tried to install Intel Compilers and MPICH2, I ran into 
big trouble.
If I use icc and ifort, when I configured MPICH2, it will complain that 
icc cannot
create C programs that run. If I use gcc instead of icc, I can make MPICH2,
but the Fortran 90 example programs cannot be made.
I looked into the error, it seems that icc and ifort created obj files 
with different
object type than the library files made from compiling MPICH2.
Due to my limited knowledge in these aspects, I finally gave up using 
Intel compilers.
I turned into using gcc, g77 and g95 as my compilers, which generate 
much slower
executables than using icc and ifort.

To have a working MPICH2 with Intel compilers, do I have to reinstall my 
system
with SL i386 version? Will >2TB disk partition be accepted by the i386 
Linux?
If both answers were 'yes', then what is the fastest way to reinstall SL 
4.[12] i386
over SL 4.1 x86_64? Do I always have to totally remove the x86_64 Linux
and install everything again?

Thanks for your advices in advance.

--
Sincerely,
Kuo Kan LIANG
Division of Mechanics,
Research Center for Applied Sciences,
Academia Sinica, Taiwan

ATOM RSS1 RSS2