Distributed mquery

It’s possible to use mquery in a distributed way:

Every agent will talk with its dedicated Ursadb instance, queries will be run on all servers in parallel and results will be merged.

In fact, the default stock configuration is really “distributed”, just with a single agent running on the same machine.

It’s also possible to do it “in reverse” - a single Ursadb instance can be connected to multiple mquery servers:

All of mquery’s core functionality works in this setup, but there are some deployment problems. For example, web interface assumes that the samples are stored (or mounted) at the same location as in the workers. If that’s not the case, it can be corrected with custom plugins.