[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [tor-relays] SENDME CELL



Sorry for the very bad Subject line. Thank you Mutt :).

David

On 30 Apr (15:59:23), David Goulet wrote:
> On 30 Apr (12:06:31), marziyeh latifi wrote:
> > Hello,
> > I have two questions about SENDME cell in TOR:
> > Is SENDME a Control cell or a Relay cell?
> 
> A SENDME cell is a control cell. See:
> 
> https://gitweb.torproject.org/torspec.git/tree/tor-spec.txt#n1531
> 
> However, SENDME exists at the stream level which affects only a specific
> stream within the circuit and thus their StreamID will be non zero which
> explains the "sometimes" in the spec.
> 
> > Which node does  interpret  SENDME cells?The node that receives them or the
> > edge node(client/exit)?
> 
> SENDMEs are always end-to-end so only the end points interpret them.
> 
> For example, when a client starts a download of a large file from an Exit
> node, the Exit will send some data (up to what we call the circuit window
> size) and wait for a SENDME cell _from_ the client to send more. In order
> word, the client sends the SENDME towards the Exit to tell it "please send me
> more data".
> 
> Now, if the client uploads data to the Exit, this is reverse where the Exit
> will send a SENDME when it is ready to receive more data.
> 
> All this is managed by the package and delivery window maintained at each end
> points. See Section 7 of torspec.txt for more details.
> 
> Cheers!
> David
> 
> -- 
> 71kN/ro+ccyP6zH5RukUX1TNXn7KjZ+E8ffp3xaYOzg=



> _______________________________________________
> tor-relays mailing list
> tor-relays@xxxxxxxxxxxxxxxxxxxx
> https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays


-- 
71kN/ro+ccyP6zH5RukUX1TNXn7KjZ+E8ffp3xaYOzg=

Attachment: signature.asc
Description: PGP signature

_______________________________________________
tor-relays mailing list
tor-relays@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays