[prev in list] [next in list] [prev in thread] [next in thread] 

List:       busybox
Subject:    Re: [PATCH] udhcpc: add support IETF RFC2131 section-4.1 retransmission strategy
From:       Rich Felker <dalias () aerifal ! cx>
Date:       2012-09-30 23:52:15
Message-ID: 20120930235215.GR254 () brightrain ! aerifal ! cx
[Download RAW message or body]

On Mon, Oct 01, 2012 at 01:20:18AM +0200, Denys Vlasenko wrote:
> On Friday 28 September 2012 09:04, Frank TEO wrote:
> > IETF RFC2131 section-4.1 retransmission strategy said
> > 
> > The client MUST adopt a retransmission strategy that incorporates a randomized \
> > exponential backoff algorithm... 
> > first retransmission SHOULD be 4 seconds ...
> > next retransmission SHOULD be 8 seconds ...
> > retransmission delay SHOULD be doubled with subsequent retransmissions up to a \
> > maximum of 64 seconds.
> 
> How is this going to be useful?
> 
> I see only one effect: after DHCP server goes up after a prolonged outage,
> clients will be slow at acquiring addresses. Not a good effect.

I agree completely. The retry interval should have an upper bound of
1-5 seconds. Maybe sysadmins want it higher, but users want it lower,
and I don't think the authors of software run by users have any
business making it disregard the user's best interests for the sake of
policies that somebody else wants.

I also don't think anybody using busybox in embedded devices would be
happy when their customers start complaining that getting back on the
network after an outage has several-minute delays...

Rich
_______________________________________________
busybox mailing list
busybox@busybox.net
http://lists.busybox.net/mailman/listinfo/busybox


[prev in list] [next in list] [prev in thread] [next in thread]