 |
| Schematic from PeerApp site. http://www.peerapp.com/Solutions/Network.aspx |
February 2, 2012
The name behind the proxy
I had an increase in visits to my blog recently due to Haaretz publication of my findings. Looking over statistics i noticed several visits from linkedin where i also have a profile. One entry have got my attention, someone from R&D at PeerApp looked at my profile. After checking their site and applying Occam's razor to all the data I'm quite sure at this point that this is exactly what BezeqInt uses as a "sneaky proxy". Also Netvision as mentioned before and Pelephone(see press release).
Makes me wonder how widespread this cache poisoning problem is. Is it a configuration problem? Is it inherent to PeerApp technology? I had an agreement with Haaretz to hold on and not publish my findings giving BezeqInt time to patch and/or reconfigure their system. Even after disclosing exact details to their technical staff it seems like nothing has being done. So far BezeqInt denied the viability of poisoning the cache. My factual tests with other people using BezeqInt lines prove otherwise.
Subscribe to:
Post Comments (Atom)
2 comments:
If you server is not configured properly to avoid caching, why are you complaining about bezeqint?
Sure it might be that your test found a real problem, but until you share the details of your test I am going to assume that you simply don't understand how caching is supposed to work in the web, especially after you said that http doesn't support caching, a statement which is totally false.
The only way you can avoid this transparent caching proxy is by not using port 80 or encrypting your traffic, because BezeqInt forcibly reroutes plain HTTP traffic. I never said that HTTP doesn't support caching. I wrote: "... HTTP protocol wasn't build with any kind of transparent proxies in mind ...". Please, read my other posts detailing how it works.
Post a Comment