[Babel-users] RTT stability inside a GRE tunnel

[Michael Richardson]

Denis Ovsienko <infrastation at yandex.ru> wrote:
    >>  This is what I would *suspect*:
    >>
    >>  Coming out of the GRE tunnel, your packet would have to be
    >> decapsulated by  the kernel, and then re-queued into the network
    >> bottom half.  Generally,  the kernel does interrupt mitigation, and
    >> since this is a soft-interrupt,  that queue is serviced from a clock
    >> interrupt.
    >>
    >>  As such, you'd see the values quantized to the clock interrupt, which
    >> is  exactly what I see in your diagram.

    > That's a worthwhile hypothesis (and not excluding ICMP
    > software-forwarded elsewhere), but shouldn't the green (ICMP in GRE)
    > RTT be strictly greater than the red (plain ICMP) RTT in this case?

I agree --- this bothered me as well.
Is the RTT being calculated by an ICMP?  I thought it was a babeld packet.

--
]               Never tell me the odds!                 | ipv6 mesh networks [
]   Michael Richardson, Sandelman Software Works        | network architect  [
]     mcr at sandelman.ca  http://www.sandelman.ca/        |   ruby on rails    [