2.1 removed the option to clean up rotated log files based on number of backups (which could be directly related to file size with max_size*max_backups)
Since we now gzip rotated log files, and have 4 separate log files, this calculation breaks down, and we can no longer rely on it to calculate disk usage.
We need to have a way to limit the total size of logs, as a safeguard to prevent a node's disk from being completely filled up.
As a bonus, we could look into removing the less important logs first (debug, info) to preserve errors and warnings.
Migrated from: https://github.com/couchbase/sync_gateway/issues/3713