0

Can you explain this log ?

142.678: [Full GC 142.678: [CMS142.764: [CMS-concurrent-mark: 0.088/0.088 secs] [Times: user=0.09 sys=0.00, real=0.09 secs] 
     (concurrent mode failure): 21888K->19536K(21888K), 0.2969878 secs] 31679K->23701K(31680K), [CMS Perm : 1489K->1489K(12288K)], 0.2972058 secs] [Times: user=0.30 sys=0.00, real=0.30 secs] 
Aravind Vel
  • 277
  • 1
  • 5
  • 18

2 Answers2

1

Your new generation is filling up so fast, the concurrent collection of the tenured generation could not make it to finish, before the new generation became full.

Try increasing new generation.

For more detailed explanation, look here.

Community
  • 1
  • 1
npe
  • 15,395
  • 1
  • 56
  • 55
1

First, you are using CMS garbage collector (find more about it at Understanding pauses of CMS collector

(concurrent mode failure) means that background garbage collector cannot reclaim memory fast enough (young collection have failed to find some free space in tenured generation to promote objects) and JVM have switched to stop-the-world FullGC.

Possible reason for that - heap is too small - young generation is too small and a lot of young objects are leaking tenured space.

Alexey Ragozin
  • 8,081
  • 1
  • 23
  • 23