cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1107
Views
65
Helpful
2
Replies

2 EPG between 2 Bridge Domain between 2 POD can't access each other

am.
Level 1
Level 1

Hi everyone,

We have some issues with our ACI connection between 2 POD. 
The 2 EPGs between 2 BD  on the same VRF that are located between 2 POD can't access each other.

 

BDVLANxxx: EPGVLANxxx : POD1 : DC
BDVLANyyy : EPGVLANyyy :  POD2 : DRC

We are using vzAny btw, so we don't need to add some contract.

 

The EP can ping to gateway but can't ping to another EPG or BD.

Any recommendations would be greatly appreciated.

2 Accepted Solutions

Accepted Solutions

Sergiu.Daniluk
VIP Alumni
VIP Alumni

Hi @am. 

Just for the sake of tshoot, create a contract with permit IP (or any) filter between the two EPGs, and test then. If it works, it means that the vzAny contract has some problems.

Anyway, the tshoot should go in this direction:

+ check if each EP can ping its own gateway

+ check if on each compute leaf you have pervasive route for each remote subnet (route pointing to spine proxy). If this is missing, then its a problem of contracts

+ check if local and remote EPs are learned in spine COOP. If not, problems might be related to: COOP/MP-BGP/EVPN

+ check if on each compute leaf you have the remote EP learned.

+ check the traffic flow in which direction is broken - ELAM Assistant can be of help here.

 

Stay safe,

Sergiu

View solution in original post

am.
Level 1
Level 1

Hi Sergiu,

Thanks for the answer,
Btw the issue has been solved, the routing issue on the Server-side.

View solution in original post

2 Replies 2

Sergiu.Daniluk
VIP Alumni
VIP Alumni

Hi @am. 

Just for the sake of tshoot, create a contract with permit IP (or any) filter between the two EPGs, and test then. If it works, it means that the vzAny contract has some problems.

Anyway, the tshoot should go in this direction:

+ check if each EP can ping its own gateway

+ check if on each compute leaf you have pervasive route for each remote subnet (route pointing to spine proxy). If this is missing, then its a problem of contracts

+ check if local and remote EPs are learned in spine COOP. If not, problems might be related to: COOP/MP-BGP/EVPN

+ check if on each compute leaf you have the remote EP learned.

+ check the traffic flow in which direction is broken - ELAM Assistant can be of help here.

 

Stay safe,

Sergiu

am.
Level 1
Level 1

Hi Sergiu,

Thanks for the answer,
Btw the issue has been solved, the routing issue on the Server-side.

Save 25% on Day-2 Operations Add-On License