Nodes recycle/restart due to a JVM memory issue. The issue is caused by high memory consumption by the Shazzam sensor in the Discovery sensor's jobs.
Steps to Reproduce
Although there are no definitive steps to reproduce the issue, the following steps might cause the problem:
- Navigate to Discovery > Discovery Schedules.
- Add the column Shazzam batch size.
- For all the schedulers, change the Shazzam batch size from 500 to 5000.
The following workarounds are available:
- Reduce the number of IP addresses in problematic schedules (split into multiple schedules)
- Do not exclude IPs from a large IP range - use multiple IP ranges instead
- Do not include a large number of small IP ranges - simplify by aggregating consecutive ranges where possible
- Reduce the number of MID servers being used by problematic schedules
- Contact Customer Support for a patch to the Shazzam sensor script
After you have applied one or more of these workarounds, turn on Discovery to test.
Related Problem: PRB660591