I've a Poweredge R620 with CentOS 6 Kernel 2.6.32-504.3.3.el6.x86_64.
It shows error messages "cpu0000 cpu2 internal error (IERR) contact support".
In the BIOS, I turn off C-State and C1E.... but the error is the same...
any other ideas?
CPU IERR are errors detected by the CPU. It is unlikely that this is an issue with the CPU. You should check the hardware/system log for errors reported around the same time of the CPU IERR. CPU IERR are usually caused by DIMMs. Check for DIMM errors on the B lane that CPU 2 controls.
Dell EMC, Enterprise Engineer
I've got similiar issue, as system is RH 6.6 with the same kernel. It's running in cluster and has HBA 6Gbps SAS card included connecting to Dell PV MD3220. We had today replace the whole server (just HDDs were taken from the old one) but the problem persists. Under heavy load we can see the following entries in /var/log/messages:
Jan 13 23:33:53 hostname kernel: mpt2sas0: _ctl_host_trace_buffer_size_show: host_trace_buffer is not registered
Jan 13 23:33:53 hostname kernel: mpt2sas0: _ctl_host_trace_buffer_show: host_trace_buffer is not registered
Jan 13 23:33:53 hostname kernel: mpt2sas0: _ctl_BRM_status_show: BRM attribute is only forwarpdrive
And the server reboots with CPU Internal Error. Any suggestions what we should check?
I've got the same issue on 2 T620 servers with dual CPU, straight after BIOS upgrade from 2.2.2 to 2.4.3, CPU 2 generates the error.
definitely not upgrading anymore until I speak to Dell and work out this issue.