Needs restart if importing buckets from a large file

[johan-bjareholt]

When I import large data the bucket list doesn't get updated and I get this error for every bucket:

[2019-03-02][08:14:13][_][ERROR] Failed to write response: Os { code: 32, kind: BrokenPipe, message: "Broken pipe" }.

But then if I restart after this all the buckets and their data is there?
Something must be wrong in the Database thread? Are mpsc requests being returned too early?

Found when testing #22

[johan-bjareholt]

I just realized that it doesn't actually need a restart, it just becomes incredibly slow.
On my laptop which has a smaller (but still pretty big) export it takes 4000-6000ms to make a /api/0/buckets request until I restart aw-server-rust and then it's instant again (3ms request).