Re: Why was oom-killer invoked?
On 3/19/2012 3:48 AM, tim truman wrote:
> Please help me understand why oom-killer was invoked?
You already know the answer: what does "oom" stand for?
>>From the SAR logs we can see that there is lots of memory in use by the
> system cache, but free is low. How can we ensure there is more memory
> available in free to avoid triggering oom-killer?
Switch to a 64 bit kernel and user space, including 64bit pgsql. This
gives you a flat memory space so any app can have more than 2GB virtual
address space. You already knew this as well, but you've been fighting
the switch tooth and nail, screaming "but I shouldn't _have to_, it
works now, mostly". It's the "mostly" part that is your current
problem. Switch to 64bit with its large flat memory space and you'll be
I still can't understand why so many people are so damn resistant to
going 64bit, *especially* database users, who benefit the most from it.