I believe that you can safely delete the .old files -- that just clears out the
history. I concur with the idea of trying it on another machine.
Mark
-Original Message-
From: [EMAIL PROTECTED] on behalf of Julian Rohrhuber
Sent: Fri 2/8/2008 8:37 AM
To: pws@cc.gatech.edu
Subject: Re: [Pws] AlwaysCrashing
maybe someone else could host the swiki for a test and see whether it
is due to the files or due to the machine / squeak image?
The situation is pretty bad - the hfbk swikis were in heavy use by a
lot of different people and I find it to be a shame that this has
been offline for such a long time now.
David T. Lewis wrote:
The swapping issue would relate to total memory usage on your
server, and may have nothing at all to do with your Squeak application.
Are you sharing the server with other people? If so, find out what
has changed, and who or what is hogging the system resources.
this box is dedicated to run the swiki only and it can only be the swiki
process hogging resources, there's nothing else running on the box.
(apart from the tiny and hardly used webserver on :80)
oh, just forgot: a LowSpaceDebugLog appeared, i put it on
http://swiki.hfbk-hamburg.de/LowSpaceDebug.log.txt
You are starting your Squeak image with a fixed object memory
allocation (-memory 128m). The low space alert presumably means
that Squeak has used up the available object memory.
I thought this could make it crash earlier, but it still took a couple
of hours. it was usually (for years) started with -headless only.
the amoount of data in the swiki is only gradually increasing.
during testing, i had two swikis not starting up, but responding with
file closed. could this mean one of them triggering the bug ? (in one
case one single swiki is 879MB, mainly for three .old pages totalling 573MB)
Just guessing,
but this may be nothing more than a side effect of the system
swapping, hence Squeak going slower, hence perhaps not able to
keep up with things in some way. You can of course increase
the fixed allocation, or remove the -memory option entirely
to permit the VM to dynamically extend object memory. Of course
this would further aggrivate the memory swapping problem, so
you need to figure out what is going on there regardless.
Dave
without the fixed allocation the same symptoms occur; i can either put
it on a P IV/1800 / 256MB or a dual PIII/1000 / 512 MB.
after five+ years, both of these machines run out of memory after some
time, while the amount of data in the swiki has meanwhile reached ~2.1 GB.
if this is another limit i'll have to ask some of the professors to
squeeze their uploads to smaller sizes. (just kidding)
other than that, i'm stumped.
cheers,
tee
btw, i'm on the list, no need to cc: me ;)
___
Pws mailing list
Pws@cc.gatech.edu
https://mailman.cc.gatech.edu/mailman/listinfo/pws
--
.
___
Pws mailing list
Pws@cc.gatech.edu
https://mailman.cc.gatech.edu/mailman/listinfo/pws
___
Pws mailing list
Pws@cc.gatech.edu
https://mailman.cc.gatech.edu/mailman/listinfo/pws
