On 9/28/2012 3:09 AM, tcr@xxxxxxxxxxxx wrote:
Hi Marcus, The 5GB is on a squid on a different server... one not running under valgrind. I've restarted that process since my last e-mail so it's using much less mem now, but be assured it will eventually creep back up to 5GB and beyond. I get the 5GB number from top and ps... they show roughly that number for both virtual and resident. It's a real problem... the machines will start swapping like crazy when they exhaust physical RAM and the servers slow down badly. Left for long enough, they'll eventually crash with a "malloc() returned null" exception or something along those lines. It's a real leak. Amos identified some leaks from my valgrind output and when those make it into a daily build I'll try again and report my findings. Thanks -Ty
I was just reading another thing you wrote and I want to ask: how many times per minute are you reconfiguring squid by using HUP? just curios? (reminds me old thing about memory leak and reconfiguring) Eliezer -- Eliezer Croitoru https://www1.ngtech.co.il IT consulting for Nonprofit organizations eliezer <at> ngtech.co.il