Faced the same issue with you. But managed to worked around the issue. The best solution is basically to upgrade the to a bigger, better machine. Use the paramedic plugin to view the utilization. I give you my example.
How I approach it: Logstash Reindex Strategy