-
MajesticBankour nodes were under heavy attack mostly coming from tor network, but also from clearnet so I made the guide what can be done to prevent unwanted traffic hitting the monerod
-
MajesticBanksuggestions are appreciated
-
MajesticBankthis is not final versions and will be improved
-
MajesticBank
-
M53a45b263745298There hasn't been much progress on the review of monero-project/monero #8774 since I opened it a month ago, is there something I can do to speed this up?
-
M53a45b263745298vtnerd: vtnerd_
-
vtnerdLooking at it now
-
M53a45b263745298You said in the PR comments "This patch mitigates but doesn't complete solve the issue". Are you talking about the issue of timing analysis or specifically circuit reuse?
-
M53a45b263745298From what I remember when writing this the only possibility of circuit reuse is when a remote peer initiates the drop because there's no socks5 to force a new circuit on next connection
-
M53a45b263745298For very high threat models using tx proxy with one connection is the best for privacy with this patch
-
M53a45b263745298Actually one or two, you want to prevent a sybil from linking connections based on when noise sending is started/stopped
-
M53a45b263745298But that makes it easier for a remote node to link you across connections based on how quickly you disconnect so it's not perfect either
-
M53a45b263745298I think the best privacy would be if the recommendation/default was one or two outgoing tx proxy connections, it means the maximum delay between successive transactions would be extended by RANDOM_CONNECTION_DROP_UPPER but would make sybil a lot less effective
-
M53a45b263745298Adding a delay to prevent instantly broadcasting after the epoch would improve privacy some more
-
plowsof11when you implement these suggestions the review can take place
-
M53a45b263745298It woul be good for somebody to confirm that there are no identifiers sent on tx proxy connections that make all of this effort useless, I couldn't find any but there was mention of "timestamp in peer timed sync"
-
M53a45b263745298Why would I implement these things before discussing?
-
M53a45b263745298I've been waiting a month to get the initial review done, I'm trying to make some progress here
-
plowsof11you are pinging a contributor for a progress update on the "review". meanwhile in the pull request i see "This is a "rough draft"/"conversation starter". I have not done much C++." so are you pinging for a review or implementation?
-
M53a45b263745298I've implemented the entire patch, these are minor adjustments
-
M53a45b263745298That's just to stress that it needs review and discussion before being merged, but it's a working patch
-
M53a45b263745298Did you read the code?
-
M53a45b263745298I'm sorry for trying to improve Monero, where would you like me to take my thoughts on doing so?