Martin Pala <address@hidden> writes:
ad. 1.) ICMP ping support - already scheduled, i will look on it. I
think i can start the implementation next week.
I think I have read someplace about a fping.c file that you can use?
If not, simply use the enclosed file. It's a ping-of-death program
with ip-spoofing, I stumbeled over a while ago and hacked a bit (nb!
this is not my work). You should probably remove the spoofing :-) and
you need to read response from the server. Combining this with net.c:
check_udp_socket() should do the trick. (To spoof replace the
xxx.xxx.xxx.xxx string with the ip-address, but to use this program as
is it is, is of course a bad idea)
ad. 3.) daemontools - it could be good to test it and add the hint to
documentation. I can do it.
Great. BTW, Klaus Alexander Seistrup sent me a notice later and told
me that my suggestion will not work:
| Actually, I think this should already be possible (I do not use
| daemontools so I cannot test it). If monit has a tty, which it
| has if it was started with the -I switch, then monit will print
| log messages *both* to the logfile/syslog and too the tty's stderr.
because isatty() is false when monit is run under daemontools. Instead
he simply used this hack in log_log:
replacing:
nolog:
if (Run.have_tty) {
with:
nolog:
if (Run.have_tty || Run.init) {
This might even be enough with a bit more code since env.c guarantee
that the std file descriptors should be open, altough they may go to
/dev/null, so a check on this is probably necessary.