MOO-cows Mailing List Archive

[Prev][Next][Index][Thread]

Re: [SERVER] USR1 signal



>I have a question... is it possible to change the meaning of the USR1
>signal, as the dump-child of the MOO process uses it ?
...
>Now if the MOO wouldn't shutdown, we could try and figure out where the
>problem lies -- in the system, the scsi-interface (swap memory), the
>physical memory, or the MOO...
>
>Does anyone know if changin the USR1 signal would be possible/useful/wise ?

For debugging, one simple tweak you could try is to comment out the code in
server.c that makes the main MOO task shut down when the child (the
checkpoint) doesn't feel like living.

Look for the line:

panic(const char *message)

...in server.c, go down a few lines, and modify the lines:

    if (in_child) {             /* We're a forked checkpointer */
        errlogf("Child shutting down parent via USR1 signal\n");
        kill(parent_pid, SIGUSR1);
        _exit(1);
    }

...to read now:

    if (in_child) {             /* We're a forked checkpointer */
        errlogf("Child NOT shutting down parent via USR1 signal\n");
      /*  kill(parent_pid, SIGUSR1); */
        _exit(1);
    }

Save, make, add salt to taste.
Good luck.

-------------------------------------------------------------
Gustavo Glusman               Founder/administrator of BioMOO
-- Gustavo@bioinformatics.weizmann.ac.il
-- http://bioinformatics.weizmann.ac.il:70/0h/Gustavo/Glusman
-- BioMOO: telnet bioinformatics.weizmann.ac.il 8888
           WWW:   http://bioinfo.weizmann.ac.il:8888




Follow-Ups:

Home | Subject Index | Thread Index