On 25/07/2013 10:01 a.m., Golden Shadow wrote:
Thanks Amos, Eliezer and Markus for your replies!
@Eliezer: The server has 2 X 2.7 GHz CPUs, each with 12 cores. Squid version is 3.3.7 compiled from source and I'm running only one squid worker.
@Marcus: What is maximum process size that TLB can address? Is it tunable? My operating system is CENTOS 6.4 with Kernel 2.6.34.
@Amos: Please see my comments below:
----- Original Message -----
From: Amos Jeffries <squid3@xxxxxxxxxxxxx>
To: squid-users@xxxxxxxxxxxxxxx
Cc:
Sent: Wednesday, July 24, 2013 9:58 PM
Subject: Re: Squid Memory and Page Faults
On 25/07/2013 1:05 a.m., Golden Shadow wrote:
Hi there!
My squid is installed on a server with 192 GB of RAM. I have the following directives in squid.conf:
cache_mem 143360 MB
maximum_object_size_in_memory 300 KB
memory_replacement_policy heap GDSF
memory_pools on
memory_pools_limit 1024 MB
ipcache_size 2048
ipcache_low 90
ipcache_high 95
fqdncache_size 2048
top reports that my squid process size is 20GB, which is far less than my RAM size,
but nevertheless I still find some page faults (about 70 page faults over 2 hours).
I'm wondering how could those page faults are occurring while squid process size is far
less than my RAM size. How can I eliminate those time consuming page faults?
Two things here.
Why is the process size only 20GB? you have a 143GB memory cache as part
of that RAM consumption by Squid. Perhapse your traffics real caching
requirement is far smaller than you are allowing storage for.
Well, according to cache manager my squid RSS reached 173 GB at some time!
Okay. That makes a lot more sense.
What exactly is the page faulting comign from though ... Squid or the OS?
How can I tell? Cache manager itself reports these page faults, so I guess they are coming from squid.
Hmm. Not sure exactly been so long since I've seen one. But IIRC it was
recorded as part of the message. eg. (timestamp, source, err message)
Markus mention of TLS might be a better clue than I'm provide though.
Amos
