argumentcomputer / argumentcomputer/zk-light-clients
Move to a more flexible hardware definition for Helm
- Dominant language
- Rust
- Stars
- 70
- Forks
- 12
- PR merge metrics
- No merged PRs in 30d
Description
The current Helm/K8S implementation for our proof server is mainly thought for a usage along the `eksctl` configuration file in the repository. It forces the users to have deployed hardware with a special tag, `r7iz.metal-32xl` to know which instances/nodes to deploy the cluster pods.
As mentioned by @huitseeker [in #228](https://github.com/argumentcomputer/zk-light-clients/pull/228#discussion_r1763186049), we could move away from this `nodeSelector` behavior and instead leverage ressources configuration.
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with the Helm/K8S implementation for the proof server and the repository's eksctl configuration file. Trace how the r7iz.metal-32xl nodeSelector is used, then determine how resource configuration should replace that requirement; done means users no longer need the special hardware tag to deploy the cluster pods.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- helm, kubernetes
- Domain
- cloud, infrastructure
- Issue type
- Refactor
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100