Transient increase in diskspace usage to 90% with increased cpu usage

Hi there - we use cloud watch and received an alarm for >90% disk usage for one data point at 92% either side was about 65%. Data points are every 10 sec. We run a dedicated ec2 instance with 50GB total disk space. Trying to understand how this could happen - any guidance appreciated. Below is a cloudwatch plot of the data points. We dont use swap space, we have a cassandra and postgres db running on it with a third party commercial platform (I have reached out to them as well). CPU also went up to about 85% from < 50%).