Damon Getsman schrieb:
Again, I'm on OpenSuSE 10.3 running SRSS 4. So I get everything
installed, being careful to leave the old version in place and easy to
switch back in case something breaks. Then I ssh in to one of the
servers in each of the clusters I administrate, and run firefox through
the ssh tunnel to my machine.
It works perfectly, verified that it is the correct version running.
Then I go to check from one of the SunRays.
Did you exit (completely) this version before you moved over to the Sun
Ray? If firefox finds another instance of firefox running on the same
machine, it typically sends any URL arguments to the running instance
instead of opening them in the new process instance (one of the reasons
for that is to prevent concurrent, interfering writes to the common
configuration data (under $HOME/.firefox). Unfortunately this happens
even if the existing instance runs on a different X display.
Nothing from the Desktop icon (which is running the same link that I
changed to point to the new version). I try it from the command line on
the SunRays. Nothing. I do an strace and get a large amount of tracing
with no anomalies and then it halts at a waitpid(-1, ...... which never
closes.
The firefox executable (e.g. /usr/bin/firefox) you stracing is a script
which does some environment setup and finally launches the real firefox
binary (typically named firefox.bin). To strace into that you either
need to edit the firefox script to do 'strace $MOZ_PROGRAM' where it
invokes $MOZ_PROGRAM near the end or use strace -f on the script (which
will bring you a lot of extra output for the initial script part ...).
That the waitpid never returns means that the firefox.bin process
doesn't terminate.
Has anybody else here attempted to install firefox 3 on their SRSS
machines for SunRay access? I really don't have a clue what the
difference in the user's runtime environment would be that would cause
the binary to just hang like that, and google certainly wouldn't have
even sent out spiders to pick up any discussion about this that may
happen after it's been out for a few more days.
Here's a block of the strace where things look like they're starting to
go bad... It's a long block and I apologize for that in advance. I'm
snipping off a large piece of it at the top that repeats at least 50
times above it, too (that'd be the first line that's repeated so much),
which was my first sign of something amiss.
You are stracing /bin/sh. I wouldn't be surprised, if that goes through
a regular routine for every line or statement parsed ....
HTH
- Jörg
_______________________________________________
SunRay-Users mailing list
[email protected]
http://www.filibeto.org/mailman/listinfo/sunray-users
