Re: Rethinking security, bursting, load management
Benjamin Caplan <[email protected]> Fri, 05 Jun 2009 16:19:05 -0500
| Newsgroups | gmane.network.freenet.technical |
|---|---|
| Message-ID | <[email protected]> |
It strikes me that bursting should be completely safe as long as the node's decision whether to burst at any given moment is based entirely on the availability of bandwidth (supply), and not at all on whether there is actually a request queued that could use it (demand). When supply exceeded demand, chaff would be sent; when demand exceeded supply, the queue would back up. This scheme might remove the entire point of bursting, though. _______________________________________________ Tech mailing list [email protected] http://emu.freenetproject.org/cgi-bin/mailman/listinfo/tech
signature.asc
(application/pgp-signature, 197 B)
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEARECAAYFAkopi84ACgkQyUdQDqgQe321hQCg3nQCIcw6eNN4sBWKxNvWc2EI yVoAoOTM6n1WgsNR9Lm4N/cIhUX6UtsQ =ztOR -----END PGP SIGNATURE-----