0

We have a number of HP EliteDesk workstations of varying models (8300, 800 G1) in our organization, some of which started exhibiting strange behavior over the past few months. Checking the Intel AMT event logs we see every 2-3 minutes the following error message being logged:

Operating system lockup or power interrupt.

The machines in question are all running under Windows 8.1.

Symptoms include system slowdown, occasional system lockups which in time become more frequent, sometimes the machines fail to boot up and just hang before the user is able to enter their login credentials (blank screen with dark blue background - not BSOD).

We have updated the BIOS on all affected machine which was a recommendation from Intel when these errors start frequently appearing in the AMT event log. It did not help - the error still keeps being logged every 2-3 minutes or so.

Also, we've updated / reinstalled device drivers, did a Windows Refresh, ran a malware scan...

We have been in contact with HP support regarding the first machine to exhibit these problems and they were unable to resolve the issues (MB, memory, processor, HDD all been replaced; finally, the entire machine was replaced with a brand new one which for now is working OK).

Now two more of our affected HP machines are starting to exhibit system lockups as well.

Question:

Since Google searches did not return much useful information, and we've tried everything we could think of to resolve these issues, we'd like to ask if anybody else experienced similar issues? How did they resolve them? Any troubleshooting suggestions, tips?

Radoye
  • 35
  • 5

2 Answers2

0

Without more detailed logs it would be hard to tell exactly what is causing the problems. One approach would be to try to boot the systems with a recovery image and attempt to extract any OS logs related to the events. The OS logs along with the AMT logs and the Server/System Event Log(SEL) and Sensor Data Repository(SDR), which can be accessed through the Intelligent Platform Management Interface(IPMI) using ipmitool or impiutil assuming the platform you are using supports it(which it probably does if you can get AMT logs).

By correlating the logs from various sources you may be able to see if there is an event that is causing the OS to crash or some sort of Out-of-Band power reset before the

Operating system lockup or power interrupt.

Log you reported. Once the root-cause of the error is identified the appropriate corrective action can be implemented.

Matt
  • 2,711
  • 1
  • 13
  • 20
  • Thank you for your input Matt, we will try looking at SEL and SDR logs as well. We have already looked at the OS logs and could not find anything that correlates to these AMT errors or that could explain the system slowdowns / freezes. However, an interesting thing that we noted while doing a Google research on this issues is that all affected users seem to be running Windows, and some among them with dual boot systems say that when they boot into Linux the issue does not seem to be happening. Which i find very interesting indeed! – Radoye Jul 10 '15 at 13:17
0

I've had similar problems on the same board.

  1. If you haven't provisioned Intel AMT yet. The problem is most likely that the autoprovisioning tries to connect to a provision-server outside your network. It's not default settings for Inte AMT and it isn't correct working so. Check your DHCP+DNS-server, and PKI-DNS field in MEBx. The solution is simple - you should do provisioning Intel AMT or disable it in BIOS.
  2. If you have provisioned Intel AMT. In this case more complicated. You can try:
    • check RAM (primarily in Slot 0/1)
    • clear CMOS
    • update BIOS (you've done it)
    • try connect your SATA-drives to other slots (in the first place - boot-drive)