I have a Java EE application running in Red Hat 6 with Java 7 and Wildfly 9. The application has been running well sometimes, suddenly the response time makes the application unusable and I have to stop Wildfly process and start again. Then the application becomes normal for a few more days. I don't have any jvm error in Wildfly server.log or any another Java error, but I took this error message with dmesg command.
NET: Registered protocol family 36
INFO: task java:14736 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
java D 0000000000000002 0 14736 1 0x00000080
ffff8804aefdddf8 0000000000000082 0000000000000000 ffff8804fb0211e8
ffffea0011529c00 ffff8804aefddde8 0000000000000000 000000011889588b
ffff8804875eba78 ffff8804aefddfd8 000000000000f598 ffff8804875eba78
Call Trace:
[<ffffffffa00e05d5>] jbd2_log_wait_commit+0xc5/0x140 [jbd2]
[<ffffffff8108e160>] ? autoremove_wake_function+0x0/0x40
[<ffffffffa00e0686>] ? __jbd2_log_start_commit+0x36/0x40 [jbd2]
[<ffffffffa00fc55c>] ext4_sync_file+0x13c/0x250 [ext4]
[<ffffffff811a09b1>] vfs_fsync_range+0xa1/0xe0
[<ffffffff811a0a5d>] vfs_fsync+0x1d/0x20
[<ffffffff811a0a9e>] do_fsync+0x3e/0x60
[<ffffffff811a0af0>] sys_fsync+0x10/0x20
[<ffffffff8100b172>] system_call_fastpath+0x16/0x1b
hrtimer: interrupt took 3019336509 ns
NET: Unregistered protocol family 36
NET: Registered protocol family 36
I have looked for an answer and applied the settings bellow into my sysctl.conf.
vm.dirty_background_ratio = 5
vm.dirty_ratio = 10
But, didn't work! Does somebody has any idea about what could cause this problem?
