[squid-users] New StoreID helper: squid_dedup

Hans-Peter Jansen hpj at urpla.net
Mon May 16 12:35:03 UTC 2016


Hi Eliezer,

Thanks for your feedback, much appreciated, /especially/ from you.

The most important part is in dedup.py. I've kept an eye on efficiency without 
sacrificing readability (much) and extendability:

	https://github.com/frispete/squid_dedup/blob/master/squid_dedup/dedup.py

A big part of the rest is related to configuration management, which tries to 
maximize convenience (as many config files as wanted, automatic reload option 
on changes, etc..)

Depending on public interest, it would be cool to create a public CDN 
collection, that is shared among users, or even distributed automatically.

Pete

On Montag, 16. Mai 2016 03:44:29 Eliezer Croitoru wrote:
> Thanks for sharing!
> 
> I didn't had enough time to understand the tool structure since I am not
> a python expert but,
> This is the first squid helper I have seen which is based on python and
> implements concurrency.
> 
> Thanks!!
> Eliezer Croitoru
> 
> On 10/05/2016 00:56, Hans-Peter Jansen wrote:
> > Hi,
> > 
> > I'm pleased to announce the availability of squid_dedup, a helper for
> > deduplicating CDN accesses, implementing the squid 3 StoreID protocol.
> > 
> > It is a multi-threaded tool, written in python3, with no further
> > dependencies, hosted at: https://github.com/frispete/squid_dedup
> > available at: https://pypi.python.org/pypi/squid-dedup
> > 
> > For openSUSE users, a ready made rpm package is available here:
> > https://build.opensuse.org/package/show/home:frispete:python3/squid_dedup
> > 
> > Any feedback is greatly appreciated.
> > 
> > Cheers,
> > Pete



More information about the squid-users mailing list