AttCron's huge RAM usage

[Chris Rende]

[ Nixdorf Targon M35/50 with TOS 3.2 ]
[ Pyramod 9810 with OSx 4.0 ]

This machine has 32Meg of RAM.
The CRON(1M) that I'm running is /etc/.attcron

I'm quite proud of the fact that it has been running for 149 days without
any problems:

$ ucb uptime

 11:38am  up 149 days, 19:21,  46 users,  load average: 0.92, 0.79, 0.86

However, yesterday I noticed this:

$ ps -efl | egrep "(NI|cron)"

  FLAGS S    UID   PID  PPID  C PRI NI     ADDR  SZ    WCHAN  STIME   TTY  TIME COMD
 108001 S   root    61     1 22  26 20     23071057 ff1108f4  Apr 22    ? 239:49 /etc/.attcron

Notice the "SZ" of CRON! It's humungous!

I decided to take a snap-shot of CRON's memory to see what was going on:

# ucb gcore 61

# ls -l core*

-rw-r--r--   1 car      user     2164736 Sep 18 16:40 /u3/car/core.61

Doing an ASCII dump of the core file showed that it is full of strings like
"/usr/tmp/a0061" and "/tmp/croutDTA" where <DTA> is a random set of upper-case
letters. That's what is taking up 2 Meg of my RAM: pages and pages of those
two strings!

The following day I dumped CRON's memory again:

# ucb gcore 61

-rw-r--r--   1 car      user     2174976 Sep 19 11:30 /u3/car/core.61

I wasn't supprised to find that CRON's memory had grown by another 10,240 bytes
overnight.

For a "work-around" I suppose that I could KILL CRON and then restart it
from time to time to free up the RAM...


Is this a bug which is fixed in newer release of Pyramid's Unix?
(The Unix I'm running is <relatively> old :-().

car.
-- 
Christopher A. Rende           Central Cartage (Nixdorf/Pyramid/SysVR2/BSD4.3)
uunet!edsews!rphroy!trux!car   Multics,DTSS,Unix,Shortwave,Scanners,UnixPC/3B1
trux!car at uunet.uu.net          Minix 1.2,PC/XT,Mac+,TRS-80 Model I,1802 ELF
trux!ramecs!car     "I don't ever remember forgetting anything." - Chris Rende