Quantcast
Channel: All Routing posts
Viewing all articles
Browse latest Browse all 8688

VPLS failing with routing-engine failover

$
0
0

Hi everybody.

 

I have a question and I am looking for a solution. PE1 and PE2 are two different locations, P3 and P4 are in the same location. I have the following topology, where PE1-PE2, PE1-P3 and PE2-P4 are connected over a DWDM waves and P3 to P4 using single mode fiber directly:

 

Drawing2.jpg

 

On PE1 and PE2, I do have 2 eBGP session for Internet Access (4 in total), each with full routing table and one iBGP session from PE1 to PE2, from PE1 to P3, from PE2 to P4 and from P3 to P4. I also have an OSPF session between from PE1 to PE2, from PE1 to P3, from PE2 to P4 and from P3 to P4.

 

Between PE1 and PE2, I do have a few VPLS tunnels. They work well and the performance is great, however, when I do run a "request chassis routing-engine release" on PE1 from the master, the BGP and OSPF sessions are transferred to the other routing engine, but the VPLS connections do fail and need up to 5 minutes to get reestablished.

 

My question is pretty simple: can I make this happen faster or better avoid any disconnection of my VPLS tunnels? If so, what should I do? I am running JunOS 15.1R6.7 on RE-S-1800x4 routing engines.

 

Any help will be highly appreciated.

 

Thanks.


Viewing all articles
Browse latest Browse all 8688

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>