Protect Journal Against Corruption

Requesting some form of protection for the journal against corruption due to runaway inflation.

When a surge of incoming logs causes the buffer to consume all available space on it's assigned volume, the journal becomes corrupted. This requires deletion of the corrupt journal and the loss of all data contained within.

## Expected Behavior

Journal should never be allowed to exceed it's maximum size setting.
## Current Behavior

Journal will allow journal to exceed max size setting.
## Possible Solution

Shorter intervals for the check, some kind of active measure when it approaches a threshold to prevent corruption from occurring.
## Steps to Reproduce (for bugs)

1.
2.
3.
4.

## Context

The larger the configured journal maximum size, the larger the impact should it become corrupted. In once case, 60GB of collected logs were lost in a single journal deletion.
## Your Environment

* Graylog Version: ALL
* Java Version:
* Elasticsearch Version:
* MongoDB Version:
* Operating System:
* Browser version: