Hi sven,
I think there's a heavy memory leak around processing macros.
process_macros_r function mallocs et reallocs many times (stored in mac structure) but these allocations are never deallocated in both handle_host_check nor handle_service_check when cleaning memory at end.
Before this update, naemon was losing about 1 Mo/minute (1000 hosts, 10000 services).
With my update, running now for 24 hours, naemon stays a 20 Mo without any leak.
Hi sven, I think there's a heavy memory leak around processing macros.
process_macros_r function mallocs et reallocs many times (stored in mac structure) but these allocations are never deallocated in both handle_host_check nor handle_service_check when cleaning memory at end.
Before this update, naemon was losing about 1 Mo/minute (1000 hosts, 10000 services). With my update, running now for 24 hours, naemon stays a 20 Mo without any leak.
What do you think about ?
jfr