0

We have moved to Java 8 few days back . We have deployed our application on physical JBOSS server on unix server. We are facing below error frequently within a week which cause those whole application getting down for a few minutes. Can anyone please advice solution for the same ?

2020-04-23T20:01:29.196+0100: 21.571: [GC (Metadata GC Threshold) [PSYoungGen: 183229K->1798K(1797120K)] 367320K->314795K(5991424K), 0.0754280 secs] [Times: user=0.67 sys=0.16, real=0.07 secs]
2020-04-23T20:01:29.272+0100: 21.647: [Full GC (Metadata GC Threshold) [PSYoungGen: 1798K->0K(1797120K)] [ParOldGen: 312996K->215034K(4194304K)] 314795K->215034K(5991424K), [Metaspace: 154812K->154780K(1191936K)], 0.3053917 secs] [Times: user=2.05 sys=0.01, real=0.31 secs]
2020-04-23T20:01:31.728+0100: 24.103: [GC (Allocation Failure) [PSYoungGen: 1520128K->155329K(1815040K)] 1735162K->370380K(6009344K), 0.0196249 secs] [Times: user=0.15 sys=0.00, real=0.02 secs]
Michael
  • 34,340
  • 9
  • 58
  • 100
Liam
  • 3
  • 1
  • 1
    Sorry, the information you have given is not sufficient for anyone to help you. Yes, there is a full GC. Yes, that means you are running out of memory. Does that mean you have a memory leak, like you tagged the question with? Maybe. We don't have any information about what's in the heap to point to what's using all that space, or any information about the code to point to a smoking gun. – Michael Apr 30 '20 at 09:43
  • 2
    _exactly_ where do you see `minutes` of application down from this log? – Eugene Apr 30 '20 at 09:43
  • 1
    I know there are people out there recommending to always do gc logging, but when the result is just having people confused about the logs and asking how to resolve an error, when there is no actual problem, the best recommendation is to turn off gc logging again. You don’t have a problem. “Allocation Failure” is not an error, but a normal condition, triggering a garbage collection. – Holger Apr 30 '20 at 09:55

0 Answers0