Hello:
I would like to thank all those people who read my email and offered me
suggestion.
Because several other people are also interested in the solution to this
problem,
I attached all the suggestions here:
Weixing Zhang
************************************************************************
************************************************************************
***
Difficult to say. Of course it is not X itself. Some browser releases
show a memory hole. To check this out, close all possibly open browser
windows. The next hot candidates are all self written programs. It's
rather common in such programs to forget the release of unused memory.
Same procedure: close that programs and have a look for the memory used
by X. Usually the operating system should release the taken resources.
************************************************************************
************************************************************************
***
http://www.bruker-biospin.com/shell/bkb/show_bug.cgi?id=5768
Item 5768
Computer freezes or slows down (Linux)
ParametersItem#: 5768
Status: closed Severity: serious
Op. System: LINUX Program: TOPSPIN Component: unknown
Version: 1.3.0 Beta Version:
Solved in: not solved ?/? Version:
Instrument: any Hardware Unit: any Hardware Group: any
Symptom: A customer has reported that the computers start slowing down
and eventually freeze completely as we run Topspin on them. The PC is HP
x1100 Linux with RedHat Enterprize 3 operating system. They tried the
following
steps:
upgraded kernel from 2.4.21-9.EL to 2.4.21-15.EL , added 512MB of RAM to
the PCs bringing total memory to 1GB upgraded video cards to 64MB.
Unfortunately these steps did not solve the problem.
After executing an Update of the operating system (Red Hat Enterprise
Linux 3.0 WS), the error did not occur again.
A possible reason for this is the following RedHat bugfix regarding
auditing:
http://rhn.redhat.com/errata/RHBA-2005-219.html
HP XW4200
Another case was reported for the HP XW 4200 PC. In this case the Update
to RHEL WS 3 Update 6 has solved the Problem. The Update modified a
NVIDIA Rel 75 driver. The old driver could cause 'X Hangs with 100% CPU
Utilization on Red Hat Enterprise Linux WS 3'.
Please check the following web page for further information:
http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c
00529204
*****************
Document ID: c00529204
Version: 1
HP xw series Workstations - X Hangs with 100% CPU Utilization on Red Hat
Enterprise Linux WS 3
NOTICE: The information in this document, including products and
software versions, is current as of the Release Date. This document is
subject to change without notice.
Release Date: 2005-10-24
Last Updated: 2005-12-06
------------------------------------------------------------------------
-------
DESCRIPTION
Occasionally X will hang with 100% CPU utilization. This may happen at
anytime X is running, but is more likely to occur when a user logs out
of X or a root user uses 'init 3' on workstations as described in the
SCOPE section.
SCOPE
HP xw series Workstations with Red Hat Enterprise Linux (RHEL) WS 3
Update 5 or earlier.
NOTE: This issue is not limited to HP systems.
RESOLUTION
Update to RHEL WS 3 Update 6.
NOTE: When using a NVIDIA Rel 75 driver (such as 76.76) the modules.conf
file must be modified to include 'options nvidia NVreg_UseCPA=1'. The
future Rel 80 series of NVIDIA drivers will not require the modules.conf
file modification.
See the NVIDIA README for information about page_attr and cache
aliasing.
------------------------------------------------------------------------
------
Hardware Platforms Affected: HP xw4200 Workstation, HP xw6200
Workstation, HP xw8200 Workstation, HP xw9300 Workstation, HP xw4300
Workstation
Components Affected: Not Applicable
Operating Systems Affected: Red Hat Enterprise Linux 3 (x86), Red Hat
Enterprise Linux 3(AMD64/EM64T) Software Affected: Not Applicable Third
Party Products Affected: Not Applicable
Support Communication Cross Reference ID: IA00524165
(c)Copyright 2005 Hewlett-Packard Development Company, L.P.
Hewlett-Packard Company shall not be liable for technical or editorial
errors or omissions contained herein. The information provided is
provided "as is"
without warranty of any kind. To the extent permitted by law, neither HP
or its affiliates, subcontractors or suppliers will be liable for
incidental, special or consequential damages including downtime cost;
lost profits; damages relating to the procurement of substitute products
or services; or damages for loss of data, or software restoration. The
information in this document is subject to change without notice.
Hewlett-Packard Company and the names of Hewlett-Packard products
referenced herein are trademarks of Hewlett- Packard Company in the
United States and other countries. Other product and company names
mentioned herein may be trademarks of their respective owners.
**********
The operating system upgrades seems to fix all these individual problems
This is an example that the Bruker BioSpin recommendation should be
accepted -
Download and install all available upgrades / hotfixes
Of course a guarantee can not be given by Bruker BioSpin that a
operating system patch will worjk correctly and does not influence
Bruker BioSpin software. But the experience shows that it is a higher
risk not to be up2date with operating system patches then to install
them.
For windows Bruker Biospin recommend to install critical updates only.
Software patches and new drivers should only be installed if it is
really necessary.
This may be related to bugs: 5762, 5380 and/or 4789
************************************************************************
************************************************************************
***
I had a similar problem when my system was first delivered running RHEL
3, but it went away when I updated the Linux.
************************************************************************
************************************************************************
***
I don't know if it is related to your problem, but a while ago we had
problem with X using almost 100% CPU, which caused Linux computer to
freeze. I have not checked memory usage at that moment. After the
update of graphics card driver the fault disappeared.
************************************************************************
************************************************************************
***
X is the graphical process manager. Do not worry about swap, unless it
is being used. Linux uses all memory (most of it is cache) and the
virtual memory system allocates lots of swap, although most does not
touch the disk.
Perhaps your X is leaking memory, control-alt-backspace kills X (and all
graphical programs!). X leaking is unlikely tho. You might check if your
system has been cracked.
************************************************************************
************************************************************************
***
I don't have any advice, but if you do learn anything, please let me
know as well. Our slow-down problem seems to be worst first thing in the
morning after the computer has been sitting most of the night.
************************************************************************
************************************************************************
***
X is the program that provides the graphical point-and-click user
interface that we are all accustomed to. On Linux, it has almost always
been a bit of a memory hog. I've never seen it slow down a computer,
but I have seen it suddenly cease interoperating with the user. Since X
(re-)starts whenever someone logs in, it is usually enough to
log-out/log-in, rather than reboot, to solve such problems. Or, you can
kill (graphical) processes until it the situation improves. If you
can't get access to a terminal window, then this can be done by a remote
login. Alternatively, you can usually hit something like a [Cnt'l]-F4
or F7, which will provide a new (non-graphical) login environment.
Exactly which keystroke to use depends on the O.S. and version.
Possible remedies include: (1) don't run too many graphical programs
(esp. web-browsing) at the same time! (b) upgrade the memory (c) upgrade
the version of X from x.org. Only (a) is guarranteed to work, (c) has a
good chance, too.
************************************************************************
************************************************************************
***
________________________________
From: Zhang, Weixing [mailto:Weixing.Zhang_at_STJUDE.ORG]
Sent: Thursday, August 30, 2007 12:59 PM
To: ammrl_at_chemnmr.colorado.edu
Subject: AMMRL: Linux computerslow down problem
Hello:
We have a Linux computer (HP XW4200) which works as the host worksattion
for a Bruker 400 MHz NMR.
ICON-NMR is used for automation. The computer works fine after it was
rebooted, but it slowed down late.
Here is the information from the file /proc/meminfo:
After the computer was rebooted, it is like this:
total: used: free: shared: buffers: cached:
Mem: 1050001408 460083200 589918208 0 23228416 187777024
Swap: 4194869248 0 4194869248
When the computer becomes too slow, it becomes:
total: used: free: shared: buffers: cached:
Mem: 1050001408 1029054464 20946944 0 3620864 764563456
Swap: 4194869248 1265819648 2929049600
A lot of memory and swap space was used.
When I used the "top" command to check processes, I found that the
command "X" run by root
uses more than 50% of the memory and it is still increasing gradually.
Do you have any suggestions to find out what might cause the slow down.
What does the "X" command do?
Thanks,
Weixing Zhang
Received on Wed Sep 05 2007 - 12:36:06 MST