[squid-users] Experimental YouTube Caching helper\tool and SQUID 3.5.25 + 4.0.19 RPM's RELEASED

Eliezer Croitoru eliezer at ngtech.co.il
Tue May 16 01:04:26 UTC 2017

Hey List,

I have been working a while ago on a tool(ICAP service) that will do the
- Snatch YouTube main video pages(watch?...) on the fly
- Fetch them
- Parse them
- Find the key to predict the ID of the X.googlevideo.com/Y which will match
the YouTube video
- Store the key with the correlated video ID in a redis DB for an hour
- Then a StoreID helper will analyze X.googlevideo.com/Y urls and will try
to see if it finds the matching a YouTube video by the ID on the redis DB

The concept is kind of "divide and conquer" while later if I will have time
there will be an option to combine these two things into one ICAP service
that will help to cache YouTube videos and probably also Windows Updates
using squid.

Currently the ICAP service code will not be published but the binaries and
the resources which I have used to create the tool will be published.
(Something like the ingredients without the recipe due to this task being
considered a caching master level and not just any novice)
However my email is wide open for questions requests and tips for anyone
that wishes to get answers, directions and guidance.

@Alex, What will encourage you to help with an improved patch to add ICAP
X-StoreID (or else) response header capability that will help me to write a
complete StoreID solution based on ICAP alone?

@Others, if you are interested in this solution publication please respond
to this thread and CC to me.


* I have released Squid-Cache 3.5.25 and 4.0.19 RPMS but Life takes more of
my time and as I get busy I didn't have enough time to publish the article
like in the past.
The article topic is "ENCRYPTION, HOW FAR WILL YOU GO?" at

Eliezer Croitoru
Linux System Administrator
Mobile: +972-5-28704261
Email: eliezer at ngtech.co.il

More information about the squid-users mailing list