CPU Usage

[beatreichenbach]

Hello,

I'm running hound from a Docker container on an alpine image based on the docker file in this repo. I have a 92 repos being indexed.
My docker container is showing medium cpu usage compared to some other containers that I have running. It's nothing bad but I'm wondering why is it constantly running at this level?

I figured if I set the ms-between-poll to 10m it wouldn't really use any cpu until it starts polling for changes. But the container is constantly sitting at that medium usage. Is there a difference between having 100 small repos indexed vs one large one?

Here's my config:

{
  "max-concurrent-indexers": 1,
  "title": "Codebase",
  "dbpath": "/hound/data",
  "vcs-config": {
    "git": {
      "ms-between-poll": "10m",
      "detect-ref": true
    }
  },
  "repos": {
    "<repo>": {
      "url": "[email protected]:<user>/<repo>.git",
      "url-pattern": {
        "base-url": "https://{url}/src/main/{path}{anchor}",
        "anchor": "#lines-{line}"
      },
      "vcs-config": {
        "ref": "main"
      }
    },
    ...
  }
}

Thank you,
Beat

[beatreichenbach]

The issue seems to be the ms-between-poll key in the config that is not supported on a global level.
Neither of these configs seem to work:

{
  "vcs-config": {
    "git": {
      "ms-between-poll": "10m"
    }
  },
  ...
}

{
  "ms-between-poll": "10m",
  ...
}

I was only able to make it work by specifying the poll time per repo, even though the docs say:

You can override this value by setting the ms-between-poll key on a per repo basis in the config.

It seems like the key is only supported on a per repo level and not on a global level.