joker helps the caching using the store_id feature of Squid 3.4 or the storeurl_rewrite feature of Squid 2.7.
It uses a plugin model, in which you can support new websites easily
without hassling too much with a long chain of if else code.
It should run on any platform that Go supports, like Linux, FreeBSD, OpenBSD, Windows, to name a few.
If you've got Go installed, you can install joker with Go's command line interface:
go get github.com/osmano807/joker
You can test your installation by running joker -version from any folder.
You can use goxc to cross compile joker to another platform.
Prebuilt packages are not offered for the moment.
I'm currently running on a OpenBSD 5.5-current machine, with the
squid-3.4.2p0 package.
On my squid.conf I have:
store_id_program /usr/local/sbin/joker -new-format
store_id_children 20 startup=0 idle=1 concurrency=100
acl joker_access_list dstdomain i.imgur.com
acl joker_access_list dstdomain .bp.blogspot.com .blogblog.com
acl joker_access_list dstdomain .glbimg.com
acl joker_access_list dstdomain .gravatar.com
store_id_access allow joker_access_list
store_id_access deny all
For now, the -new-format is required to enable the store_id
format, if not, it will be compatible with the storeurl_rewrite
feature. It supports concurrency, but I don't have a huge load of
traffic to test the optimal options.
- Profile the code and identify any spots for improvement.
- Create a system to each plugin output his required squid acl.
- Create more tests on the code.
This program is inspired by the inComum project, a similar project.
Also, inspired by the project to create a packaged squid for Windows and the works on the store_id feature on Squid 3-HEAD.
The code is released under the BSD 2-Clause License.
I'm coding to distract myself from problems, or, for fun. I don't have much time to support the code or to help, but feel free to mail me, I'll try to give an reply as soon as possible.