Good day. Im currently deploying mx routers with rpms and event-options configured for multiple ISPs with "change commands" inorder to do a route failover.
I enabled load-balance per prefix on the MX router.
Majority of the sites have 2 isps. I have more than 100 sites with juniper firewalls facing the MX.
Having 100 sites with probes and event-options could be resource intensive on the MX part. What is the best practice to do this?
So each site has MX router connected to 2 ISP?
And Why do you need to configure ~100 rpms and event option on each router? You can only configure 2 rpm, one for each ISP to do a route failover on each MX.
mx is seated in my DR who monitors atleast 60 branch srx devices.
mx is configured to have a dual ISPs with load balance per prefix (default routes per isp)
mx will monitor each isp router interface facing the srx.
the srx is configured with dual isps also (load balance per packet ---> flow is packet-mode)
im thinkin if creating a slax script for the monitoring or simply event-options with probes to trigger a route failover.
the only problem with this setup, is im not able to make it work ...-> when either isp1 or isp2 detected a link failure on their monitored (rpm probe) ips, route didn't failover.
the MX continuously prefer the route towards the branch with a disconnected / failed interface.
© 1999 - 2019 Juniper Networks, Inc.
All rights reserved