[Dibbler] Issue with short lifetimes
kauer at biplane.com.au
Mon May 24 02:29:27 CEST 2010
On Sun, 2010-05-23 at 14:58 +0200, Tomasz Mrugalski wrote:
> What values were sent by the server? I assume client behaved according
> to T1 and T2 values sent by server, right?
Don't know. I did NOT set T1 and T2 in the server, but I DID set
lifetimes. The client then used inappropriate defaults OR the server
then sent inappropriate defaults.
> Actually, this should be fixed on server side. I think (please confirm
> that this indeed is the case) you set lifetimes, but didn't explicitly
> set T1 and T2 in server.conf, so server provided short prefered and
> valid lifetime, but large T1 and T2. Client behaved as instructed by
Correct. I did NOT set T1 and T2 in the server, but I DID set lifetimes.
I do not know if the server sent any T1 and T2 values, I did not check
However, I disagree that the fault should be fixed only in the server.
The client should reject bad T1 and T2 values, I would define "bad" as
any value equal to or greater than the preferred lifetime. If bad values
are received, it should use sensible defaults and issue a warning in the
The server should IMHO reject a configuration where T1 or T2 are equal
to or greater than the configured preferred lifetime. It should either
reject the configuration as broken or should adjust T1 and T2 to
sensible values and issue a warning in the log.
In the meantime, the workaround is to make sure that the server is
configured with T1 and T2 values that are appropriate to the configured
> But your point is valid - that should be fixed. I'll fix that shortly
> after my Ph.D defense. According to the latest forecasts, this
> cataclysmic event is expected to happen in about a month.
Karl Auer (kauer at biplane.com.au) +61-2-64957160 (h)
http://www.biplane.com.au/~kauer/ +61-428-957160 (mob)
GPG fingerprint: B386 7819 B227 2961 8301 C5A9 2EBC 754B CD97 0156
Old fingerprint: 07F3 1DF9 9D45 8BCD 7DD5 00CE 4A44 6A03 F43A 7DEF
More information about the Dibbler