Hi Matt, On Wednesday, 9. June 2010, Jan P. Kessler wrote: > Am 03.06.2010 21:29, schrieb Jan Wagner: > > On Tuesday 25 May 2010, Matt Taggart wrote: > >> I notice that postfwd 2 is available now. It looks like the rulesets are > >> compatiable so hopefully it's a drop in replacement. Do you know > >> > >> * how it compares stability/performance wise to v1? > > > > is still the only different the preforking server and use of > > Net::Server::Multiplex? > > postfwd2 still does not fully support ratelimits. Limits do only work > within a single child process. > > Except this the difference is only the architecture. postfwd1 is a > single multiplexer, while postfwd2 consists of three subsystems: a > master 8control) daemon, a cache server (multiplexer) and the policy > server (prefork). If you are interested you'll find some information at > http://www.heinlein-support.de/upload/mk4/2-03_postfwd-Das-flexible-Policyd > -Framework.pdf, page 31ff. > > The ruleset parser is the same code and both versions take the same > command line arguments. If you ask me, I'd consider postfwd2 as the > 'more robust' version. Due to the prefork model and the builtin watchdog > function, a single request can never cause the whole function to > terminate. I you postfwd2 on different medium to large scale production > systems. > > > Do you think, it makes sense provide v1 and v2 in Debian? Actually we > > just provide v1. > > Yes, if it is not too much work for you it might make sense. postfwd1 > has ratelimits and needs fewer ressources (at my home I run postfwd on a > NAS box). Especially when using dns based stuff on large scale systems > postfwd2 is the prefered choice.
from the answer of Jan P., I would say replacing postfwd v1 shouldn't happen (yet). Beside that postfwd2 hasn't a upstream tarball, at least I haven't found one. Do you feel, we need postfwd2 in squeeze? I personly wouldn't hammer postfwd2 into squeeze, cause the testing period would be really short. With kind regards, Jan. -- Never write mail to <w...@spamfalle.info>, you have been warned! -----BEGIN GEEK CODE BLOCK----- Version: 3.12 GIT d-- s+: a C+++ UL++++ P+ L+++ E--- W+++ N+++ o++ K++ w--- O M V- PS PE Y++ PGP++ t-- 5 X R tv- b+ DI D+ G++ e++ h---- r+++ y++++ ------END GEEK CODE BLOCK------
signature.asc
Description: This is a digitally signed message part.