Support #12032
closed
  
ip-90-147-166-112.ct1.garrservices.it: too many opened connection to node66.p.d4science
 
        
        Added by Roberto Cirillo over 7 years ago.
        Updated over 7 years ago.
        
Infrastructure:
Production
 
 
  
  
  
  Description
  
  Who is this host?
I'm not able to access to this host: ip-90-147-166-112.ct1.garrservices.it 
Please could you give me access? We should understand why there are a lot of connections opened from this host to the following host:
node66.p.d4science.research-infrastructures.eu:8080
   
 
  
  Files
  
 
  
  
    
    
    
    
       - Target version changed from UnSprintable to node66.p: too many open connection
 
    
    
     
   
  
  
    
    
    
    
       - Related to Task #12031: gCubeApps - IS-Collector: too many open connections added
 
    
    
     
   
  
  
    
    
    
    
       - Assignee changed from _InfraScience Systems Engineer to Roberto Cirillo
 
    
    It's the IP address that's used by the hosts without a public IP. So there's one (or more than one) of the private IP dataminer that went crazy.
 
     
   
  
  
    
    
    
    Andrea Dell'Amico wrote:
It's the IP address that's used by the hosts without a public IP. So there's one (or more than one) of the private IP dataminer that went crazy.
Could I have access to this VM? Otherwise the next month I'll ask you the same question: who is this host? 
Maybe yes, there is a crazy dataminer behind this host. I need further analysis.
 
     
   
  
  
    
    
    
    
    I've verified the logs on all the dataminer instances with private ip.  I see the same exceptions found on ckan-bb1, at the time when the IS-Collector was restarted.  In attachment the log of dm-192-168-100-6 instance.
Maybe it's better to try to restart these containers in order to improve (I hope) the IS-Collector performance. What do you think @andrea.dellamico@isti.cnr.it ?
 
     
   
  
  
    
    
    
    Go on. One at a time, you can restart each one.
If the connections problem persists, we can even restart all the dataminers once a day. If we restart the service only when no jobs are running, there's no risk of breaking anything.
 
     
   
  
  
    
    
    
    
       - Status changed from Feedback to In Progress
 
       - Assignee changed from Andrea Dell'Amico to Roberto Cirillo
 
    
    OK. I'm going to restart the services one at a time
 
     
   
  
  
    
    
    
    
       - Status changed from In Progress to Closed
 
       - % Done changed from 0 to 100
 
    
    The services have been restarted. Now I see 0 opened connections from 90.147.166.112. I'm going to close this ticket.
 
     
   
  
 
  
  
 
Also available in:  Atom
  PDF