Hello,

We are having an issue and have not yet identified the root cause. The issue is actually effecting Oracle RAC driving wait states so high that a node is dropped. In looking at Linux to see if this is derived from an OS issue I notice this happening before the wait states issue occurs:

Code:
Jul 25 14:54:40 mlblpdnaorav01 kernel: INFO: task ocssd.bin:5600 blocked for more than 120 seconds.
Jul 25 14:54:40 mlblpdnaorav01 kernel: "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
Jul 25 14:54:40 mlblpdnaorav01 kernel: ocssd.bin       D 0000000000000000     0  5600      1 0x00000080
Jul 25 14:54:40 mlblpdnaorav01 kernel: ffff881435e1f980 0000000000000082 ffff881fe6ddae00 ffff881435e1ffd8
Jul 25 14:54:40 mlblpdnaorav01 kernel: ffff881435e1ffd8 ffff881435e1ffd8 ffff881fe6ddae00 ffff880fffc14780
Jul 25 14:54:40 mlblpdnaorav01 kernel: 0000000000000000 7fffffffffffffff 0000000000000000 0000000000000000
Jul 25 14:54:40 mlblpdnaorav01 kernel: Call Trace:
All things being equal, would this be indicative of the Linux OS itself causing the hung service or Linux simply reporting an Oracle hung service?

Thanks for helping me interpret this,


Michael