Reduce pipeline v2 caching

[astrsh]

Pull Request

Brief description.

Reduces the amount of cached pipeline chunks from 1024 to 64, and reduces the target size of chunks from 500 to 128. This represents an approximate reduction from a 16k x 16k block area to a more sensible 1024 x 1024 block area per pipeline provider loaded by the server.

What Issues Does This Fix?

Fixes excessive memory usage, reported by @ccorp2002 on discord.

Licensing

• I am the original author of this code, and I am willing to release it
  under GPLv3.
• I am not the original author of this code, but it is in public domain or
  released under GPLv3 or a
  compatible license.

Types of changes

• Bug Fix
• Build system
• Documentation
• New Feature
• Performance
• Refactoring
• Repository
• Revert
• Style
• Tests
• Translation

Compatiblity

• Breaking
  change
• Non-Breaking change.

Contribution Guidelines.

• I have read
  the CONTRIBUTING.md
  document in the root of the git repository.
• My code follows the code style for this
  project.

Documentation

• My change requires a change to the documentation.
• I have updated the documentation accordingly.

Testing

• I have added tests to cover my changes.
• All new and existing tests passed.

[solonovamax]

has it been measured what the original memory usage was from this, and how much this causes it to decrease by?

Also, this should be a config option

[astrsh]

From C_Corp's report on discord, there was ~12gb worth of PipelineBiome[] instances, with a decrease down to 9gb by simply reducing cache size by some amount (not sure what exactly he decreased it by) - usage will vary by how many packs are installed but it probably shouldn't be that high.

Wasn't too scientific with picking numbers, just figured a lower default would be better for now, then revisit it if and or when it becomes an issue (unlikely it will be until more packs that use pipeline v2 such as OW2 are more prevalent), or memory usage is critically examined. Cache size could probably even be smaller, cache hits I believe only tend to happen in a fairly local area, a larger cache would still be somewhat beneficial for the speed of /locate I would think but this would need to be profiled.

Already opened #402 regarding it being a config option.

[solonovamax]

yeah, this definitely needs to be profiled further & cache hits/misses need to be recorded to see how large we should try and make it for optimal performance/memory tradeoff