Steps to reproduce:

1. No  methord to reproduce this bug

Result:

1. crash log below , the first node(active node ) crash 

   3.After the crash, the first node is start by himself and became a standy node

  4. But 20 minutes later, history poller is very busy and cache usage increase and then decrease

Expected:
1. What is the cause of the zabbix server crash and is there already a corresponding fix for it

2. Why history poller  High utilization