Re: [tor-talk] IDEA: Compress traffic at exit

[grarpamp <grarpamp@xxxxxxxxx>]

> We decided at the time that it was ok to abandon it on the theory that a)
> most big things on the internet are compressed already, and b) if it's
> not a big thing, then compressing it isn't really going to buy you a
> whole lot. It would be great to see some numbers to support or refute
> this theory.

People could run netflow logging on their exits and easily rank
things by port utilization. Then when you see TCP/80 up there
in the top 10, you can dump off an hour's worth to disk and
test the compression ratio there. Since Tor is port agnostic
internally, just dump everything going in the directions you
care about. It's rough due to it be a layer 1 instead of application
layer tap. Still, uncompressibles stick out as near 1500 byte
packets of just that.

This is science, not the tapping card, so don't trot it out.

> Theory: Good webservers compress by default if the browser headers
> indicate it's supported; and good browsers offer to receive compressed
> content by default.
>
> Again, it would be great to get some numbers to support/refute. It would
> also be good to confirm that we don't screw any of that up with Torbutton.

I don't know where to point you for such an application layer test. That's
an if we can, is it being done question. Any vendor of ISP cache hardware
will happily spam you with their observations about the net :) Or fire off a
note to nanog.

> Ultimately, the deal-breaker here is that the legal liability question
> shifts dramatically when you change from being a transit provider ("bytes
> go in, bytes go out") to having content stored "at" the relays. From a
> technical perspective we might just put the cache in ram and thus there's
> no stored content; but it's really not worth introducing the possibility
> that some poor exit relay operator will have to sit down with the judge,
> jury, and lawyers and try to teach them how computers work -- in terms
> that fit a telephone metaphor since that's what most of the laws expect.

I don't see a distinction between ram, disk, linecard or wherever.
Just like 'storing' [voice]mail. So long as it is being done agnostically
as an ISP would, it would hardly be an indefensible issue. That's like
saying gmail is responsible for all the gerbil pron scattered amongst
it's disks. Of course if one starts caching only gerbilpron.com, yeah,
that might be kinda hard to explain as being a hands off approach.
They came to you because of the traffic. If you can explain
that, you're in pretty good shape. If not, well... . In my jurisdiction,
gerbil pron is highly frowned upon, so even with disk encryption
or ramdrives, I might be inclined not to do it unless at a datacenter.
_______________________________________________
tor-talk mailing list
tor-talk@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-talk