Improvements to the Web3 anonymization service

Hey All!

As part of the update released today, we’ve been having huge improvements to the ⋮⋮⋮ Web Anonymization apparatus. How so? We’ve largely improved the ⋮⋮⋮ Decentralized Cache mechanics. Now, every 6 hours (meaning 4 times a day), it would be made sure that the cached data remains in sync with the target website.

Previously we’ve been seeing stale data (even for days or weeks) all too often.

While cache data in the first place? To improve performance and efficiency of incoming client requests. So that ⋮⋮⋮ Nodes do not need to query remote endpoints every time a request arrives for the same resource on the Internet.

Assets cached by the service, make their way out of the system on LRU( Least Recently Used) basis, meaning that the content which is accessed least often disappears the soonest.

What is important to note, dynamic content (such as videos) but also including images - are never cached and are queried from the remote endpoint on per-access basis.

Additional benefits of the caching service include the now much lowered overall CPU utilization, since contents need to be mangled with only once.

Do you mind explaining redundancy properties of the ⋮⋮⋮ Decentralized Cache ?

Well, well it comes to redundancy properties, as of now,- each node is ‘on its own.’

I do not see any imminent be altering that.

Let me explain.

Each node maintains a cache of Web data on its own. No dynamic content is ever cached.
There is no synchronization mechanics of cache data between nodes. I do not see any benefits to introduce such, at least - as of now. I say it even had the benefit if improved decentralization guarantees.

Then, when user chooses to use the ⋮⋮⋮ Decentralized Cache indirectly, through the ⋮⋮⋮ Browser UI dApp, the ⋮⋮⋮ Browser UI dApp would toss a node serving data on a per-request basis. @PauliX please confirm.

this idea looks great

[image]