Chemolli Francesco (USI) wrote:
> I'm using the good old printf, as suggested by Adrian.
> Since this is run on a semi-production server (it's the only way to get
> the Nxxthousand hits to trigger the bug), having squid crash is more
> acceptable than just stopping responding (if squid hangs but is still
> bound to the port, PAC-based failover won't work).
What I commonly did in the same situation I installed a couple of GDB
scripts to automate the data collection when the fault was detected, but
if you are in a position that you can get coredumps, printf+abort() is
just as good, except that it requires a recompile to change the data
collection setup.
-- HenrikReceived on Tue Jun 05 2001 - 10:34:42 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:14:03 MST