blume@netmbx.UUCP (Heiko Blume) (03/28/89)
i recently caused some bad load plus angry root on this machine
(286 xenix 2.2.1) because i compiled the wrong thing by chance.
it was the common process well
main() {
for (;;) fork();
}
which, of course, caused the no more processes message. i thought
'no problem', hung up the phone and redialed. the beasts were still
running ! and it was impossible to kill them because kill seems
to be too slow (one syscall for each pid) to that effect that the
forkers had enough time to spawn again before all of em got the signal.
the machine had to be rebooted.
the question is: why did they survive the logout ? they were not nohup.
(don't try this if you can't reboot, it's a real pain for the machine).
perhaps they could have been killed if the kill would run at top priority ?
--
Heiko Blume,Seekorso 29,D-1000 Berlin 22,VOICE=(+49 30)365 55 71,BBS=()365 75 01
TELEX=183003 intro d,FAX=()882 50 65|ARPA =crash!pnet01!pro-cess!blume@nosc.mil
PSI =PSI%45300043109::netmbx:blume |BITNET=pro-cess.UUCP!blume@PSUVAX1
UUCP =blume@netmbx.UUCP |INET =blume@pro-cess.cts.comtim@cunixc.cc.columbia.edu (Timothy Jones) (03/28/89)
Doing a "kill -STOP" for each running process will freeze the programs (not allow them to regenerate). Then, when they are all frozen, you can "kill -9" them successfully. ****************************************************************************** Timothy Jones Columbia University Center for Computing Activities Instructional Computing Consulting Group BITNET: tim@cunixc, timus@cuvmb ARPANET: tim@cunixc.cc.columbia.edu USENET: ...!rutgers!columbia!cunixc.cc.columbia.edu!tim ******************************************************************************
dyer@spdcc.COM (Steve Dyer) (03/29/89)
In article <1346@cunixc.cc.columbia.edu> tim@cunixc.cc.columbia.edu (Timothy Jones) writes: >Doing a "kill -STOP" for each running process will freeze the programs >(not allow them to regenerate). Then, when they are all frozen, you can >"kill -9" them successfully. Say, what? Did job control sneak into XENIX 386 or UNIX V.3/386? :-) :-) -- Steve Dyer dyer@ursa-major.spdcc.com aka {ima,harvard,rayssd,linus,m2c}!spdcc!dyer dyer@arktouros.mit.edu
raulin@tdl.UUCP (Raulin Olivera) (03/30/89)
In article <2464@netmbx.UUCP>, blume@netmbx.UUCP (Heiko Blume) writes: > i recently caused some bad load plus angry root on this machine > (286 xenix 2.2.1) because i compiled the wrong thing by chance. > it was the common process well > main() { > for (;;) fork(); > } > which, of course, caused the no more processes message. i thought > 'no problem', hung up the phone and redialed. the beasts were still > running ! and it was impossible to kill them because kill seems > to be too slow (one syscall for each pid) to that effect that the > forkers had enough time to spawn again before all of em got the signal. > the machine had to be rebooted. > the question is: why did they survive the logout ? they were not nohup. > (don't try this if you can't reboot, it's a real pain for the machine). > perhaps they could have been killed if the kill would run at top priority ? Did you try useing a kill 0? It is supposed to kill everything that you initiated. You might add the nice to raise the priority. As far as not dying I suggest that you check the HUPCL settings in gettydefs. =Ralo->