Add xxx-padding-machines-for-onion-clients.txt. #72
Conversation
| out. In particular, the origin-side machine needs in average 3.3 secs to | ||
| deplete all its padding cells, and hence it will slightly increase the | ||
| circuit lifetime. See section [IMPROVEMENTS] for a further low-hanging-fruit | ||
| improvement in this direction. |
Where is the IMPROVEMENTS section?
Also, the manage_circ_lifetime field should keep the circuit open for MaxCircuitDirtiness seconds now. Is that not the case?
Oops, the IMPROVEMENTS section was removed. It was suggesting to split the machine into multiple phases, so that we try to hide the post-circuit-construction traffic as well.
Good point that the #28780 also messes with MaxCircuitDiriness need to update that.
|
|
||
| XXX Isn't this kinda intense????? Using the graphs from metrics we see that | ||
| the Tor network has total capacity of 300 Gbit/s which is about 135000GB per | ||
| hour, so 3.9GB per hour is not that much, but still... |
Nah, this is fine
We need to skate to where the puck is going to be, or we'll deadlock forever on stuff like this and never deploy anything. The plan is to implement proposal #265 and ask people to run more middle relays, at which point 3.9GB per hour will become negligible (it's already close to that, since exits are often scarce by more than that much anyway).
ACK
No description provided.
The text was updated successfully, but these errors were encountered: