Skip to content
This repository has been archived by the owner on Mar 11, 2021. It is now read-only.

Remove cluster cache implementation #750

Open
dipak-pawar opened this issue Dec 21, 2018 · 3 comments
Open

Remove cluster cache implementation #750

dipak-pawar opened this issue Dec 21, 2018 · 3 comments

Comments

@dipak-pawar
Copy link
Contributor

External clients (such as tenant service, oso-proxy, jenkins-idler, etc) are proxied to cluster service. The cache is not used.
Internal clients (different auth services, like account linking) should not ask for the list of all clusters at all. They should ask for the particular cluster with particular URL. And they should call our cluster service every time.

@dipak-pawar
Copy link
Contributor Author

@alexeykazakov I was thinking even internal clients needs a particular cluster, I believe if we have cache then we don't need to call cluster service each time as we can get required cluster info from cache which will definitely good for performance

If we remove cache, each time internal clients have to call cluster service for any single request which will degrade the performance.

Do you see any other good reason to remove cache?

@alexeykazakov
Copy link
Contributor

Internal clients is basically token linking stuff. It’s not used that often. Proper cache invalidation is not a trivial task and can cause issues. Let’s start with a simple solution (call cluster service when we need cluster info). We can improve it later when we face performance issues.

@dipak-pawar
Copy link
Contributor Author

Currently, we don't have any API in cluster service which will give you single cluster using URL. I will add it first in cluster service and use it in auth, because there is no point to get all clusters each time and filter it later.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants