Hi Bruce,
I've seen this before when the /tmp directory was full. I guess the problem
also occurs when /tmp becomes unwritable.
In RHEL4 and alike the filesystem is remounted read-only if some I/O error
(like a SCSI error) is encountered. Check /proc/mounts to see which
filesystems are read-only. And have a look at /var/log/messages for the
reason.
For fixing this: check what's wrong with the disk. If it's SATA, maybe you
have a weak cable or connector. Or maybe the disk really goes down the
drain. You can try checking the filesystem (as it's RO, this should not be
dangerous). But I must warn you, I once lost a filesystem when I later
remounted it RW (there were real I/O errors). So better is to check it, reboot
and watch whether you get the problem again.
Regards,
Erich
On Thursday 24 November 2005 09:54, Bruce Becker wrote:
> Hello OSCAR friends....
>
> I have been getting for the past couple of days strange output on one
> of my nodes, when issuing cexec :
> default-manpath/1.0.1(45):ERROR:102: Tcl command execution failed: if
> { [file exists /etc/man.config] } {
> set manconfig [open "|egrep ^MANPATH /etc/man.config" "r"]
> while { [eof $manconfig] == 0 } {
> gets $manconfig line
> set words [split $line]
>
> # To be blunt, I didn't have the time or inclination to figure out the
> # right TCL syntax to get the regexp right, above -- all I got was
> # "starting the line with MANPATH", but there's still a few items
> # in /etc/man.config that can start with MANPATH, but not *be* MANPATH
> # (e.g., "MANPATH_MAP"). So just double check here that we got
> # MANPATH, and not anything else.
>
> if { [lindex $words 0] == "MANPATH" } {
> append-path MANPATH [lindex $words 1]
> }
> }
> close $manconfig
> }
>
> switcher/1.0.13(91):ERROR:102: Tcl command execution failed: if {
> $have_switcher && ! $am_removing } {
> process_switcher_output "announce" [exec switcher --announce]
>
> # Now invoke the switcher perl script to get a list of the modules
> # that need to be loaded. If we get a non-empty string back, load
> # them. Only do this if we're loading the module.
>
> process_switcher_output "load" [exec switcher --show-exec]
> }
>
> On further investiagtion, the problem seems to be only coincidentally
> related to cexec, but seems to be a disk problem. We had a disk
> failure on this node before and had to reboot after doing an fsck
> which ran fine.
>
> Logging into the node as root, I get the following message
> Last login: Fri Nov 11 11:02:57 2005 from sandiego.phy.uct.ac.za
> /usr/X11R6/bin/xauth: error in locking authority file /root/.Xauthority
> init.c(375):ERROR:50: Cannot open file '' for 'append'
> chmod: changing permissions of `/root/.ssh/authorized_keys': Read-only
> file system
> chmod: changing permissions of `/root/.ssh/authorized_keys2':
> Read-only file system
>
> So, I'm more convinced that it's a filesystem problem.
>
> Has anyone seen this before ? What do the symptoms mean ? And most
> importantly, how to fix them ?
>
> Thanks very much
> Bruce
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
_______________________________________________
Oscar-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/oscar-users
