Spectrum Monitoring of the SystemEDGE Agent produces constant or occasional Management Agent Lost alerts in spectrum events console, polling resumes but the problem re-occurs.
SystemEDGE has 2 modes Polling, and Listen/Respond to SNMP requests. Depending on the size of the SystemEDGE monitored system, and the responsiveness of system calls as well as other Variables (network and other). Snmp request latency (time to respond) may be slower. This is especially more common when using the SystemEDGE plugin VCAIM (Vcenter AIM) which tends to do a lot of work.
1. Update the Spectrum Models DCM_TIMEOUT value to be a higher value default is 3000 (3 seconds) updating this to 6000 (6 seconds many times alleviates this).
2. Is there an Large ACL on the SystemEDGE side? If so you may need to contact support many Versions of SystemEDGE across various platforms have fixes to change the ACL lookup function.