Hello Manju,
agent.metrics.limit has a different property in apm-events-thresholds-config.xml file. There are multiple fixes for this.
1) You need to increase this clamp limit if you dont want to see this but this is not advisable
2) You can do smartstor cleanup but again that rogue agent will keep on sending too many metrics and smartstor will fill it again
3) The real fix is to find that agent which is sending too many metrics and have to work on agent fine-tuning. Use metrics count and EM logs to find the problematic agent.
I had worked with Dan for few scenarios on this so he must understand this in a better way. Harish can also help you on this.
Thanks,
Karthik