Affects Version/s: 4.5, 4.5.1
Fix Version/s: None
Environment:HQ server 18.104.22.168-EE-9 running on CentOS 5.4
HQ-agent 22.214.171.124-EE-9 running on CentOS 5.4
- Gemfire 6.5 on hq-agent
Gemfire plugin - should remove services from inventory that are no longer available.
When stopping a Gemfire CacheServer or Gateway, Hyperic will detect the server or gateway has been stopped and report unavailable and no metrics. After restarting the stopped server or gateway, a new name is generated with it's new PID. Hyperic detects the new name as a new server or gateway. The previous server or gateway name becomes unavailable.
We should remove those services that are no longer available.
Main focus should be availability and metrics are reported for current services without showing too many unavailable services that are no longer available
After numerous stops and restarts, the list of services grows to show more unavailble services than are available
Steps to Reproduce:
1. Add Gemfire Distributed System to Hyperic
2. Allow servers and services to be discovered
3. Allow metrics to collect and report
4. Stop a server or gateway on Gemfire
5. Allow time for Hyperic to detect server status change
6. Look at the Dashboard
7. Note the cache server will show as unavailable and no metrics
8. Restart the server or gateway on Gemfire
9. Allow time for Hyperic to detect server status change or manually force auto-discovery
10. Select the Gemfire Distributed System platform
11. Note the restarted server or gateway is displayed as new name
12. Repeat steps 4 through 11
13. Note list of services can grow to show more unavailable than actual available