object_bytes() -> PANIC

• Date: Sun, 26 May 1996 07:18:46 PDT
• From:
  Richard Godard <janus@cam.org>
• Content-Type: text/plain; charset="us-ascii"

UNIX(r) System V Release 4.0 (tecfasun1)
[12]tecfasun1% uname -a
SunOS tecfasun1 5.4 Generic_101945-32 sun4m sparc

from top:
Memory: 84M real, 4008K free, 136M swap, 232M free swap

The server is the LambdaMOO server version 1.8.0p5, the only extension is a
pem_decode() bf that load in extensions.c (just to say there is no change
in the core of the server.)

May 26 07:59:48: CHECKPOINTING on TECFAMOO.db.new.#328# finished
May 26 08:20:05: *** PANIC: Caught signal 11
May 26 08:20:05: #3602:object_size object_bytes get_object_size
set_object_size (this == #213), line 2:  server panic
May 26 08:20:05: ... called from #2871:measure_object, line 22
May 26 08:20:05: ... called from #2871:measurement_task, line 9
May 26 08:20:05: ... called from #2871:measurement_task, line 15
May 26 08:20:05: (End of traceback)
May 26 08:20:05: PANIC-DUMPING on TECFAMOO.db.new.PANIC ...
May 26 08:20:05: PANIC-DUMPING: Writing 3643 objects...
May 26 08:20:22: PANIC-DUMPING: Done writing 3643 objects...
May 26 08:20:22: PANIC-DUMPING: Writing 5106 MOO verb programs...
May 26 08:20:40: PANIC-DUMPING: Done writing 5000 verb programs...
May 26 08:20:40: PANIC-DUMPING: Done writing 5106 verb programs...
May 26 08:20:40: PANIC-DUMPING: Writing forked and suspended tasks...
May 26 08:20:40: PANIC-DUMPING: Writing list of formerly active connections...
May 26 08:20:40: PANIC-DUMPING on TECFAMOO.db.new.PANIC finished

So... I restart:

>@list #213:object_size 2
#$# /* Object ~Gato does not define that verb, but its ancestor $object
does. */
@program $object:object_size   this none this
 2:  new_size = object_bytes(this);

>@d #213
Gato (#213) [ (player) programmer ]
  Owned by Gato (#213).

Let's see if object_bytes() is bogus...

>;object_bytes(#213)
=> 3976

Nope... let's check the measuement task...

>;#2871:measurement_task()
=> 0
[used 35913 ticks, 2 seconds.]

Works fine...

That's weird! Does anyone has a clue of what it could be? (it seems the
machine has plenty of free swap and my bet is that on Sun morning not much
people were busy doing anything in the MOO/on the machine)

Thanks in advance.

Janus

• Prev: Re: Not understanding eval traceback (Seconded)
• Next: Odd shutdown.
• Index: MOO-cows
• Thread: MOO-cows