Hello,
I have a scenario with two Apigee Hybrid GKE-based clusters. They are in the same region, same zone, same project and are sharing same VPC. What I want is to keep those separated in terms of proxies, products, apps, devs and such, still using the same project. Is this even possible and have manageable solution like this?
I don't have Cassandra synced as like I mentioned I want to keep data separated and currently getting data inconsistencies whenever checking UI or making a request to googleapis to get products for instance - one time response payload is empty, another I get the results; same goes for UI (error stating "Products were not loaded successfully. Error: API Product [test] does not exist for tenant [tenant-name] and id [null]").
How can I ensure such errors will not happen?
BR
Solved! Go to Solution.
One question before I give you my opinion:
Are these GKE clusters part of the same Apigee Organization or different? Looking at the title of the question, I am guessing they are part of the same Apigee Org.
In that case, you can't choose to keep different data in the Cassandra cluster of the respective GKE clusters.
If the two GKE clusters belong to the same Apigee Organization, they have to be set up in a way where the Cassandra clusters communicate with each other and continuously replicate data between each other. More details here: https://cloud.google.com/apigee/docs/hybrid/v1.7/multi-region
But, if you want to keep your proxies in different clusters, then what you can do is have different environments in those GKE clusters (you can achieve this with the overrides.yaml file), in that case one of your environments (and hence the proxies) will be running in one GKE cluster and the other one in the second GKE cluster. Even so, the Cassandra will have data of both the environments and it will be the same in both the GKE clusters.
One question before I give you my opinion:
Are these GKE clusters part of the same Apigee Organization or different? Looking at the title of the question, I am guessing they are part of the same Apigee Org.
In that case, you can't choose to keep different data in the Cassandra cluster of the respective GKE clusters.
If the two GKE clusters belong to the same Apigee Organization, they have to be set up in a way where the Cassandra clusters communicate with each other and continuously replicate data between each other. More details here: https://cloud.google.com/apigee/docs/hybrid/v1.7/multi-region
But, if you want to keep your proxies in different clusters, then what you can do is have different environments in those GKE clusters (you can achieve this with the overrides.yaml file), in that case one of your environments (and hence the proxies) will be running in one GKE cluster and the other one in the second GKE cluster. Even so, the Cassandra will have data of both the environments and it will be the same in both the GKE clusters.