Bug in System V relating to accounting
Robert N. Berlinger
Consider the following scenario:
1. User 'a' logs out at 10pm. A new getty is spawned for the
tty and hangs around waiting for next lucky soul to come
along.
2. User 'b' comes in at 9am the next morning and logs in to the
same tty. He logs out five minutes later.
3. User 'b' comes to me the next morning. He's noticed that he
has some non-prime cpu charges, even though he never logged in
at night.
I did a little research and found the problem. The "start time"
of his shell was 10pm, the night before, and therefore got
charged for the cpu of his shell as if he was on all night (as
far as p/np goes). This is because 'exec' does not reset the
start time to the current time, and user b's shell is the same
process as the getty that was started the night before.
This is somewhat of a philosophical problem. Should the start
time be the creation time of the process (i.e., pid), or should
it be the time of the most recent 'exec'?
In any case, it's a bug in System V accounting, and the same
problem might exist in 4.2BSD.
Suggested fixes (other than major mods to getty to make it fork,
I can't even think of the ramifications of that one...) would be
welcome.
--
Robert Berlinger
...{philabs,cucard,pegasus,ihnp4,rocky2}!aecom!naftoli
nar...@wg3b20.uucp
Just use the -t flag to getty in the inittab file
eg.
02:2:respawn:/etc/getty -t60 tty02 9600
This forces a new getty every minute on that line.
It is a not a very clean work around. But it would
do the trick.
--
Narayan Mohanram
Phone: 415-962 7170
ARPANET wg3b20!nar...@lll-tis-b.ARPA
Usenet ihnp4!{amdahl,pesnta}!wg3b20!narayan
Mail The Wollongong Group
1129 San Antonio Road
Palo Alto, CA 94303. USA
=========================================================
|| If you can't lick it, try some whipped cream ||
=========================================================