[squid-users] squid cache Large rock with aufs optimization for bandwidth saving

[Amos Jeffries]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 17/11/2014 2:24 a.m., Ahmed Allzaeem wrote:
> Hi Amos , thank you
> 
> Really sorry to ask you  , wts needed to be done to use largerock
> support ?

Example 5GB rock cache;

 cache_dir rock /ssd1 5000

> 
> I mean wt wring im doing ?
> 
> Also about the cace_mem directive , as I understood this value  per
> worker or per process. And here im not using rock , so its not
> shared , I have about 32 Gram total.

http://www.squid-cache.org/Doc/config/memory_cache_shared/

You are using workers, so it is shared unless you configure:
  memory_cache_shared off

Note this: "Currently, entities exceeding 32KB in size cannot be shared.".
 I am not sure if that was fixed when large-rock support was added.
Should have been, but maybe not.


For most of the per-worker settings you have are undone by the global
settings at the end of the config between http_port and WCCP config.


> 
> Can you correct nme about setting cache_mem  directive ? also guide
> me wts needed to enable caching large object size like 512k - 1 M
> or even 2 M
> 

Try "memory_cache_shared off" at the top of your config.

Amos
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (MingW32)

iQEcBAEBAgAGBQJUaD7vAAoJELJo5wb/XPRjDfYH/1SsTIG4RUrU5NXB7C4Vojva
iZdt9hPLW7QTkUWjjB3tr6Y84I4DRCd9/N6h1LBlDY3lNmjhsYQEwCEUmm6Dm2Uo
hj1P1Nvn6cBMOjDZswFrz3imo37UAnvIUAqqwMC2SJF3uglvgtrd8IdeFVPRIoSR
bgjsiZB1yAx+SBCEOPs2sEhpqdi0215FeiF/ujTS5A7JeEPS+jPVcZoYT/xRoe9L
oDXkITcfBlOa/0bY1lo5LAqPHmVDh5BZxVn8ToMMvki21J9UscRdvT04JVoc6iqP
Uouo9VxZJDJfxF0BtK4wJSG1nKtcTiFYrBqK862L2c7zwYPB9WIPpUflKRx/aFQ=
=82cL
-----END PGP SIGNATURE-----