Increased disk usage 0.16 -> 0.18

[valid orbit]

We recently updated Dagger and have been seeing some of our Jenkins workers hit disk full issues since.

Has anything changed on your end that would explain this? Has anything to the garbage collection config changed upstream in BuildKit?

Haven't dug in yet, thought I'd start by asking and seeing if you have any pointers on figuring out if/how much Dagger is contributing to this.

[static crystal]

what did you upgrade from?

and to?

query {
  engine {
    localCache {
      entrySet {
        entries {
          description
          diskSpaceBytes
          createdTimeUnixNano
          activelyUsed
        }
      }
    }
  }
}

if you're interested in getting a list of all the things using the cache, the above query can help work out what's using all the space