02-26-2014 04:21 AM
Hi
I have a question about ASR9K nvEdge cluster.
Normally,when the EOBC & IRL link are disconnected, the cluster will enter into "split node" status.
each rack will think itself is primary DSC.
My question is that after the EOBC & IRL are reconneted(cluster rebuild), whick one (rack0 or rack1) will be the primary DSC? and why?
Thanks,
Tom
Solved! Go to Solution.
02-26-2014 08:41 PM
Yes that's perfect thanks for the details. This is the expected behavior. In fact this is how we will be achieving nv ssu. Which is a new feature on the roadmap. What happens here is rack0 goes down and it sends out beacons while booting on eobc(your step 5) finds that there is a responding DSC (Rack1) gets the sw from rack 1 and reboots and joins the cluster as nonDSC.
Have you had a chance to review the nv edge deployment guide? It has these details.
Eddie.
Sent from Cisco Technical Support iPhone App
02-26-2014 08:47 AM
Hi Tom,
By design Rack0 will be preferred as primary DSC.
~Chander
02-26-2014 03:19 PM
Hi Chander
Thanks for response,
I have tried that scenario at lab, YES, Rack0 will win the primary DSC as you say.
But I tried another scenario, and it make me confuse.
I reload the Rack0 BEFORE rebuilding nVEdge cluster, Rack1 win the primary DSC.
Is it correct?
How does Rack1 detect that Rack0 is reloaded while EOBC & IRL are disconnected at the given time?
Tom
Sent from Cisco Technical Support iPhone App
02-26-2014 07:31 PM
Walk me through the steps from the beginning and we will be able to explain the reason.
Eddie.
Sent from Cisco Technical Support iPhone App
02-26-2014 07:53 PM
Hi Eddie
1. Check cluster status(Rack0 is primary DSC, Rack1 is backup DSC)
2.separate cluster by disconnect EOBC & IRL
3. Rack0 & Rack1 think itself is primary DSC
4. Reload Rack0
5. Re-connect EOBC
6. Rack0 reload again
7. Rack0 join cluster successfully
8. Re-connect IRL
9. Now Rack1 is primary DSC , Rack0 becomes backup DSC.
Sent from Cisco Technical Support iPhone App
02-26-2014 08:41 PM
Yes that's perfect thanks for the details. This is the expected behavior. In fact this is how we will be achieving nv ssu. Which is a new feature on the roadmap. What happens here is rack0 goes down and it sends out beacons while booting on eobc(your step 5) finds that there is a responding DSC (Rack1) gets the sw from rack 1 and reboots and joins the cluster as nonDSC.
Have you had a chance to review the nv edge deployment guide? It has these details.
Eddie.
Sent from Cisco Technical Support iPhone App
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide