Prometheus too many restarts
warning

Description Prometheus has restarted more than twice in the last 15 minutes. It might be crashlooping.
Query
>>>
	
				
					
					
						changes
					
				
			
				
					(
				
			
				
					
				
			
				
					{job=~"prometheus|pushgateway|alertmanager"}[15m]) > 2
				
			
    
Query Explanation

This alert fires if any Prometheus, Pushgateway, or Alertmanager instance has restarted more than twice within the last 15 minutes. The changes() function counts how many times the process_start_time_seconds metric (which indicates when a process started) has changed, signaling a restart.

Get Alert
Download
Copy to Clipboard