Kapacitor 1.6.2 will support auto-creating the task logs database

Auto-create task logs database

https://github.com/influxdata/kapacitor/commit/3e5944a1b05e2469d25401400306918b194954fd#diff-dded5a29bd00d7167799de608b0a8b18cef0d92e6c6c1a73b83a5e6d11733098R41

Default bucket/database name is "kapacitor_fluxtask_logs" . New `./kapacitord config` will print this and if the empty string is given as a bucket name `kapacitor_fluxtask_logs` is used instead.

If the database specified does not exist in the linked InfluxDB, Kapacitor will attempt to create it (with `CREATE DATABASE foo` or `POST /api/v2/buckets`). This may fail and block Kapacitor startup if the bucket does not exist and the token / user Kapacitor has is not permissioned to create buckets.

Any bucket name can be auto-created, not just the default. But if a 1.x bucket + retention policy is given like `task-run-bucket=mydb/myrp` Kapacitor will not try to create it (this feature may be added in the future but is not currently implemented).

Note the existing docs should also change in that a 1.x database with default retention policy can be given as `task-run-bucket=mydb` rather than `task-run-bucket=mydb/`.

##### Relevant URLs
Existing docs: https://docs.influxdata.com/kapacitor/v1.6/working/flux/
Issue: https://github.com/influxdata/kapacitor/issues/2620