Machine Health Screen

The Machine Health Screen shows the status of the most important components of a GGC machine, on the connected monitor.

If Google detects a problem with a GGC machine we may ask you to share a picture of what you see on the machine's monitor. This helps us resolve machine related issues more quickly and with fewer on-site visits.

When to use the Machine Health Screen

Some common cases where the Machine Health Screen can help with problem diagnosis:

• You need to reliably identify a machine that you need to work on. The Machine Health Screen displays hostname, service tag and IP address of the GGC machine.
• We ask you to investigate and reboot a machine that’s unreachable. The Machine Health Screen shows if the machine has network connectivity, and provides some information on the cause of the outage.
• We’ve detected that a hardware component has failed, or our hardware vendor has sent you a replacement part to swap. The Machine Health Screen shows you which component has failed, so you can swap in the new part correctly, then it shows if the replacement part is healthy.
• Our hardware vendor has carried out on-site repairs, and wants to check machine health before leaving your site.

How to use the Machine Health Screen

1. Connect a monitor to a VGA port on the machine. “Machine Health Screen” should be displayed on the monitor.
2. If you don't see the Machine Health Screen, it may be because you’re still installing GGC software on the machine and the GGC installer software is using the screen of the machine. To switch to the Machine Health Screen, press ALT+F2. Press ALT+F1 to switch back to the GGC Installer.
3. Check if the countdown timer at the bottom of the screen is changing. If it isn’t, the machine may have crashed and needs to be rebooted.
4. If you see the details shown below and no errors are reported, the machine is online. If not, take a picture of the screen and send it to [e-mail].

Available information

At the top of the screen you see the machine's hostname, IP address, hardware model, service tag and time since the machine was last rebooted.

At the bottom of the screen you see a countdown timer. When the countdown reaches zero, the screen is automatically refreshed.

It's important to make sure the timer is counting down. If it isn't, it's possible the machine has crashed and needs to be rebooted.

In the middle of screen are several sections, described below.

Disk status

If all disks detected by the machine are healthy, this section shows an OK message in green.

If there are failed disks, the slot numbers of failed disks is shown. This allows you to identify and replace the failed disks, and verify that the replacement was successful.

Power supply status

Two power supply units (PSUs) are installed in all GGC machines. Both should be connected to live power feeds. If both PSUs are healthy and connected to live feeds, you see a green OK status message.

If either PSU is disconnected, reports an error, or has failed, you see a red error message and information about which power supplies are problematic.

If you see “BMC unresponsive”, we’re having problems reading data from the PSUs. The machine needs to be cold rebooted. It’s likely that we've already requested that you perform this procedure.

Fan status

Multiple fans are installed in the machine. If all fans are healthy, you see a green OK status message.

If any of the fans have failed, you see a red error message and information on which fans are problematic.

If you see “BMC unresponsive”, we’re having problems reading data from the fans. The machine needs to be cold rebooted. It's likely that we've already requested that you perform this procedure.

Memory status

Here you will find the size of memory detected by the machine. This section does not report errors. This information is useful if we request that you re-seat memory modules, and will help verify if the size of memory has changed after maintenance.

Network interface status

This section shows the current status of all physical network interfaces on the GGC machine. This allows you to diagnose physical connectivity issues, and determine if the problem is with the machine, network switch, or cabling.

For each network interface, you will see:

• Interface name
• Physical connection speed in Gbps
• Interface status: UP/DOWN

Network connectivity

This section shows if the machine is able to reach destinations that are on your local network, and that are outside of it. These may include:

• The machine's default gateway
• Google DNS servers
• ipv6.google.com (for IPv6-only GGC nodes)

If all destinations are reachable, you will see a green OK status.