On a related note, we've been using InfluxDB and running into an issue where the instance will occasionally have writes take several orders of magnitude longer

On a related note, we've been using InfluxDB and running into an issue where the instance will occasionally have writes take several orders of magnitude longer than expected (to the point where a single write can take upwards of a second). Has anyone else encountered behavior like this?