"Manoj Kumar Reddy" <> wrote in message
news: om...
> Hi Guys,
>
> i got a questions regarding loadbalancing on Cisco routers. say, I
> have two locations( SiteA and SiteB) connected by two E1(LinkA, LinkB
> of equal cost) Links and i am using load balancing on these two links.
> i am using VoIP between Site A and Site B,so voice packets are also
> routed via these two links. its everything ok if these two links are
> stable. what happens when one of the links(say Link A) is
> inconsistent.
Basically you are now running the most sensitive error detection protocol in
common use on a packet network (voip) - consistant packet loss, topology
changes etc will cause quality problems with voip before any conventional
data traffic problems show up.
I mean if the LinkA is going up and down very frequently
> ( say for every 30 seconds), what happens to the voice packets?can the
> packets that are routed on Link A reach their destination?
No - you have a "black hole" - packets sent that way are never seen again.
i suppose,
> if the packets are lost on link A, i can't ask the speaker to
> reproduce the packets like how data packets are regenerated by their
> applications, if they are lost. I think IOS maintains the route
> corresponding to failed link for some period of time in its routing
> table.
EIGRP runs recalcs on an event, but with a delay to dampen the number of
changes found
what happenes to the voice packets routed via this particular
> link A in the time between the link failure and the time of deletion
> of route by IOS from its routing table?
Gone.
You may want to look at LQM which can measure the reliability of a link and
take it out of service if it falls below a threshold.
http://www.cisco.com/en/US/products/...730a.html#3911
>
> how the IOS behaves in this situation and how the packets are routed
> to destination and still using loadbalancing and EIGRP? how can I do
> this?
I dont think EIGRP is the issue - the same stuff happens with any routing
protocol, in that some kinds of faults depend on a timeout, so there is a
delay between a fault and detection of a fault.
>
> any one plz help me.
>
> regards
> Manoj.
--
Regards
Stephen Hope - remove xx from email to reply