Hey all, I wanted to gauge interest in constructing a page (possibly something that lives under https://helix.apache.org/*-docs/index.html) that explains the various Apache Helix configuration parameters including their scopes (e.g. CLUSTER/RESOURCE/INSTANCE), their default values and descriptions. I ended up constructing a first pass for my own consumption and figured it might be useful to other Helix users as well.
I don't mind generating a first draft and sharing it (though I do need to get approval for that first), but I'll very likely be missing information or be wrong on certain things so it will definitely require proofreading from more knowledgeable folks (i.e. you all). It would be cool if we could auto-generate it from the code so it didn't need to be updated separately, but I'm not sure how feasible that is at the moment so this seemed like the next best bet. This is more of a Zookeeper-level description at the moment; it doesn't currently reference the corresponding Java objects/methods. A very minimal sample formatted in Markdown is below (pasted into your Markdown previewer of choice). Thoughts? | Param Name | Scope | Deprecated | Default Value | Description | | ----------- | ----------- | ---------- | ------------- | ----------- | | DELAY_REBALANCE_TIME | CLUSTER | No | -1 | Delayed time that Helix should hold off for until rebalancing (in milliseconds). Only valid when DELAY_REBALANCE_ENABLED is true. | |...| ... | ... | ... | ... | | MAX_PARTITIONS_PER_INSTANCE | RESOURCE | No | Integer.MAX_VALUE | The maximum number of partitions, for the given resource, that can be placed on any single instance. | |...| ... | ... | ... | ... | | DOMAIN | INSTANCE | No | null | Domain represents a hierarchy identifier for an instance. The value should mirror the TOPOLOGY setting for the cluster. See https://engineering.linkedin.com/blog/2017/07/powering-helix_s-auto-rebalancer-with-topology-aware-partition-p. |
