Juniper's pig-headedness about it's 'ping' implementation has been a
pain from the start. Arguing with them gets you nowhere as they say
that they are doing it correctly while Cisco is broken.
I will not choose to argue with either, but I will say that the Cisco
behavior is functional as it allows you to test from a router to a
remote loopback. On a Juniper you never leave the box when you ping an
interface. This is of ZERO value to me.
So, while Juniper's ping may be "correct" in this, it is of no
use. While Cisco's may be wrong, it is very valuable in
troubleshooting. Regardless of compliance with the RFCs, I think
Juniper is broken.
I've spoken with many people at Juniper who agree with me, but they
all tell me that the way to do (or don't do) pings is mandated at the
senior engineering level of Juniper and has not been open to review.
Pig-headed, anal-retentive senior engineer at work...sounds like
something I might do. :-)
In any case, I would love to see lots of really BIG customers beat on
Juniper about this. Maybe they will get the point!
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman@es.net Phone: +1 510 486-8634
This archive was generated by hypermail 2b29 : Mon Aug 05 2002 - 10:42:41 EDT