ISP Link Redundancy - Load Sharing / Gateway keeps traffic in a single ISP

[KrivU]

Hi,

I have a cluster configuration, on this, we recently changed the configuration for the ISP Link Redundancy from Primary/BackUp to Load Sharing, the SMS and Gateways are on 81.10 take 99. with two gateways, one called Primary and the other is the BackUp, the ISP Link Redundacy handles three diferent ISP's.

When the traffic is going through the BackUp the ISP Link Redundacy in Load Sharing, works perfectly, we did some fw monitors and the traffic balance between the three ISP's, but if the traffic goes through the Primary-GW sends all the traffic to only one of the three ISP, We tried to run an fw isp_link {name of the ISP} down, and ran a fw monitor, and keeps sending the traffic from the same ISP,  

Has anyone else got experience a similiar behavior?

[the_rock]

Can you check the routes on both firewalls to make sure its correct?

Andy

[KrivU]

Thanks for the reply,

Yes, I checked, and the default routes are the same, are pointing to the main ISP, what other items shall I look over?

[the_rock]

Maybe run ip r g 8.8.8.8 from expert mode to confirm its right output.

Andy

[KrivU]

Hi the_rock, 

Yes I tried on both of the members of the cluster, when I failover the traffic to the BackUp, and I did the fw monitor, I can see the traffic swiching between ISPs interfaces, but when I failover back the traffic to the primary, in that cluster member the traffic just stays on one of the 3 ISPs

[the_rock]

Did TAC give you any other suggestions?

Andy

[KrivU]

No, the case still open, and they are testing in lab environment that's the last update

[the_rock]

Please keep us posted.

Regards,

Andy

[Chris_Atkinson]

Does the issue persists across reboots of that member?

[the_rock]

Excellent point Chris.

[KrivU]

Hi Chris, thanks for the reply,

I havent tried a reboot, we the TAC we tried a CPSTOP, CPSTART, but nothing changed, I'm going to schedule a reboot, to see the behavior,

[the_rock]

Hey @KrivU 

Were you able to figure this out?

Best,

Andy

[KrivU · Accepted solution]

Hi @the_rock 

We proceed with a reboot of the members of the cluster, and everything is working fine, I've been monitoring the behavior and its working as it suppose to, thanks for the follow up, the TAC told me that could be a configuration on the ISP stickiness on the firewall, but after reboot it wasn't necessary to change anything.

[the_rock]

Awesome job!