rpc.metad process consuming 25% of CPU
I have more than one V490 with 14x.35GHz processors running Solaris 10 1/06 which have the rpc.metad process consuming 25%+ of the cpu. I haved used truss and dtrace and learned that there is no disk i/o going on. The cpu utilization by the kernel is normal. This server has several filesystems mirrored across the 2 internal disks. No errors are reported via metastat, dmesg, metadb, or in the system logs.
It appears the rpc.metad process is caught in a loop (reading stdin?). Here is the output from truss which repeat and repeats...
17642: fstat(0, 0xFFBFF8C8)= 0
17642: ioctl(0, I_PEEK, 0xFFBFF9A4)= 1
17642: lwp_sigmask(SIG_SETMASK, 0x00000000, 0x00000000) = 0xFFBFFEFF [0x0000FFFF]
17642: pollsys(0x00034C40, 1, 0x00000000, 0x00000000) = 1
17642: lwp_sigmask(SIG_SETMASK, 0x00002000, 0x00000000) = oxFFBFFEFF [0x0000FFFF]
Any thoughts or known issues?
Thank you.
