route-map RIP into OSPF redistribution creates routing loops
Posted: 21 Oct 2019 10:37
Hello all,
Is any other command on Alcatel to suppress RIP route updates for new subnets?
I have found this but it seems it doesn't work for new advertised subnets.
ip rip force-holddowntimer
("Configures the forced hold-down timer value, in seconds, that defines an amount of time during which
routing information regarding better paths is suppressed")
We try to offer resiliency for a 6850 core switch running RIP and OSPF local switches connected to it over a RIP 6450 switch 1 Gb link.
The 6450 switch is connected directly to the Core 6850 switch that is redistributing RIP into OSPF and is having the uplink over 10Gb to another 10K switch which is redistributing RIP into OSPF as well.
The switches are connected like this: 6450_1Gb_RIP_to_6850_RIP_into_OSPF_10Gb_to_10K_RIP_into_OSPF_connected back to _10Gb_6450_RIP
I can attach a diagram to clarify our design.
The issue is, sometimes, when one side gets the RIP updates before the other side/switch it creates a routing loop for locally connected subnets on 6850 as it redistributing RIP into OSPF and 6850 switch listen to OSPF. I know it sound confusing, it is all in my head, is all about route maps, redistribution and avoiding routing loops.
I would prefer not to change the metric for route-map RIP into OSPF redistribution on core switches as it would affect other things.
Thank you for your time.
Is any other command on Alcatel to suppress RIP route updates for new subnets?
I have found this but it seems it doesn't work for new advertised subnets.
ip rip force-holddowntimer
("Configures the forced hold-down timer value, in seconds, that defines an amount of time during which
routing information regarding better paths is suppressed")
We try to offer resiliency for a 6850 core switch running RIP and OSPF local switches connected to it over a RIP 6450 switch 1 Gb link.
The 6450 switch is connected directly to the Core 6850 switch that is redistributing RIP into OSPF and is having the uplink over 10Gb to another 10K switch which is redistributing RIP into OSPF as well.
The switches are connected like this: 6450_1Gb_RIP_to_6850_RIP_into_OSPF_10Gb_to_10K_RIP_into_OSPF_connected back to _10Gb_6450_RIP
I can attach a diagram to clarify our design.
The issue is, sometimes, when one side gets the RIP updates before the other side/switch it creates a routing loop for locally connected subnets on 6850 as it redistributing RIP into OSPF and 6850 switch listen to OSPF. I know it sound confusing, it is all in my head, is all about route maps, redistribution and avoiding routing loops.
I would prefer not to change the metric for route-map RIP into OSPF redistribution on core switches as it would affect other things.
Thank you for your time.
