ignite-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Vladislav Pyatkov (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (IGNITE-4747) Memory leak on massive cache operations over atomic cache
Date Wed, 22 Feb 2017 16:17:44 GMT

     [ https://issues.apache.org/jira/browse/IGNITE-4747?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Vladislav Pyatkov updated IGNITE-4747:
--------------------------------------
    Description: 
When starts several nodes (I have tested in 4 nodes) with the application, through some time
(depends of heap size), we have got a Full GC pause and segmentation the grid.
After heap dump analysis, I found some suspicious object:

  Class|| Instance Count|| Total Size 
class java.util.concurrent.ConcurrentSkipListMap$Node| 63622411|           1526937864  
class java.util.concurrent.ConcurrentSkipListMap$Index|  31810595|           763454280  
class java.lang.Long|                                                         63622318|  
         508978544  


the leak in the fils "updates" of class GridDhtLocalPartition.

  was:
When starts several nodes (I have tested in 4 nodes) with the application, through some time
(depends of heap size), we have got a Full GC pause and segmentation the grid.
After heap dump analysis, I found some suspicious object:
{noformat}
  Class|| Instance Count|| Total Size 
class java.util.concurrent.ConcurrentSkipListMap$Node| 63622411|           1526937864  
class java.util.concurrent.ConcurrentSkipListMap$Index|  31810595|           763454280  
class java.lang.Long|                                                         63622318|  
         508978544  
{noformat}

the leak in the fils "updates" of class GridDhtLocalPartition.


> Memory leak on massive cache operations over atomic cache
> ---------------------------------------------------------
>
>                 Key: IGNITE-4747
>                 URL: https://issues.apache.org/jira/browse/IGNITE-4747
>             Project: Ignite
>          Issue Type: Bug
>            Reporter: Vladislav Pyatkov
>
> When starts several nodes (I have tested in 4 nodes) with the application, through some
time (depends of heap size), we have got a Full GC pause and segmentation the grid.
> After heap dump analysis, I found some suspicious object:
>   Class|| Instance Count|| Total Size 
> class java.util.concurrent.ConcurrentSkipListMap$Node| 63622411|           1526937864
 
> class java.util.concurrent.ConcurrentSkipListMap$Index|  31810595|           763454280
 
> class java.lang.Long|                                                         63622318|
           508978544  
> the leak in the fils "updates" of class GridDhtLocalPartition.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message