All things related to multicast

[Bill Owens <>]

At 10:55 -0800 2/7/01, Hugh LaMaster wrote:
> How widespread were the Juniper problems?  What version of
> JunOS was running?  (I think 4.2R2.4 is fairly current.) 
> Were the Juniper problems on boxes that had many MSDP peers? 
> What patches are desired?  (e.g. the SA-limit stuff like Cisco has?) 
> What changes would restore everyone's comfort level?

I had a single episode of serious problems with one of my two M20s, 
the one located in Buffalo. That was running JUNOS 4.0R1.2 with an IP 
I SSB and logged RPD_SCHED_SLIP errors that led to OSPF adjacency 
losses and BGP sessions dropping. However, that only happened during 
one storm period, the one on 02/01 from about 0056 - 0221 (UTC). My 
other M20, with 4.2R2.4 and an IP II SSB, never logged any errors. 
The Buffalo router has since been upgraded so that the two are twins, 
and neither had problems with the storm this morning.

I hesistate to make sweeping statements about MSDP since I've become 
re-involved with multicast deployment just within the last two 
months, after an abscence of a couple of years. But it does seem like 
MSDP is a somewhat problematic protocol, and the design changes that 
are being considered for it in the working group will require another 
round of patch upgrades for the infrastructure. And it was always my 
understanding that MSDP was a stopgap until BGMP was deployed and 
address assignment/claim issues were resolved. What's the likelihood 
of that happening, and MSDP going away? Is that the right goal to 
drive towards?

Bill.