Strange behaviour of mail11 on Ultrix/RISC

[Jens Trier Rasmussen]

Hej

We are using the DECnet-Ultrix/Internet GW for e-mail quite heavily and
have discovered a strange problem.

We can have a lot of mail11 processes in the following states (where some
of the processes are more than 12 hours old):

Ty User     Proc# Command
co daemon    2175 mail11 
    |        2176 ** Exit **
co daemon    3063 mail11 
    |        3064 ** Exit **
co daemon    6274 mail11 
    |        6275 ** Exit **
co daemon   11560 mail11 
    |       11561 ** Exit **
co daemon   12138 mail11 
    |       12139 ** Exit **
co daemon   16612 mail11 
    |       16613 ** Exit **
co daemon   20072 mail11 
    |       20074 ** Exit **
co daemon   20562 mail11 
    |       20563 ** Exit **
co daemon   21881 mail11 
    |       21882 ** Exit **
93 (16108k) processes, 2 (1276k) busy, 30 (3520k) loaded, 54 (3628k) swapped

Such couple of processes have the following status:

Ty User     Status Fl Nice Virtual Resident %M  Time Child %C Proc# Command
co daemon   kmem           104+ 96    0+  0  0   0.4        0 21881 mail11 
   daemon   EXIT                                              21882 ** Exit **
94 (16388k) processes, 1 (376k) busy, 31 (3756k) loaded, 54 (3732k) swapped

Both these displays were obtained with the sps program and we are running
Ultrix/RISC version 3.1 with the virtual memory patches from DEC. And it
is a DECsystem 3100 with 12 Mb of memory.

Have you any ideas ?

cheers

	Jens Trier Rasmussen
	CERN-CN/CS
	jenst at cernvax.cern.ch