[cisco-voip] Extended Ping Anomaly

[Jim Reed]

Here is the scenario.  I have a 2851 router in location A that is both a
voice router and a Sprint dual T1 MPLS data WAN router.

In location B, I have a 3845 router that services a few PtP T1s and a Sprint
9-Meg frac DS3 MPLS.  Location B has a separate 2851 voice router.

If I do an extended ping from the 3845 router at location B to the 2851
router at location A, I get a significant number of request timeouts.

If I do an extended ping from the 2851 router at location B to the 2851
router at location A, I don't get a single request timeout.

These extended pings have been tried with different data patterns and I
usually run five (5) sets that step through the sizes from 36 to 18024 for a
total of about 90,000+ pings.

I can also ping location A from a 3845 router at another location that is
running a Sprint 6 x T1 MPLS without a single request timeout.

This is a bizarre behavior that I have never seen before.  And it's even
more perplexing because the pings from the 2851 router at location B have to
run through the MPLS connection on the 3845 router at location B.

Unfortunately, most of the resources that location A accesses are located at
location B and are accessed via the Sprint MPLS network.  At first I
suspected Sprint but then last night I just decided for grin and giggles to
run the ping test from their voice router and the damn thing worked
flawlessly even though the pings had to go from the 2851 through the 3845.

Anyone got any thoughts.

Thank You...
-- 
Jim Reed
Manager of Technical Services
Swift Communications, Inc.
970-384-9141 (Direct)
775-772-7666 (Cell)
Sorry, no faxes accepted.
Please send documents by eMail.