Oracle Virtual Compute Appliance X4-2 Hardware
 
Oracle Virtual Compute Appliance X3-2 Hardware
 
Private Cloud Appliance X5-2 Hardware
 

PLA-Support>Sun Systems>x86>Engineered Systems HW>SN-x86: OVCA
 

In this Document

Goal

Solution

References

Applies to:

Goal

The Oracle PCA controller software contains a monitoring service, which is started and stopped with the ovca service on the active management node
The appliance administrator can retrieve current component health status information through the Oracle PCA CLI at any time by means of the diagnose command.

Solution

This document shows steps to check the health status on the PCA.

   #1 How to check the health status on the compute nodes and the zfs heads
   #2 How to identify the location of the faulted node
   #3 How to check the health status on the two management nodes
Optional steps if needed
   #4 An ILOM snapshot from the faulted node may be needed for additional troubleshooting.
   #5 How to find additional information for any faults found

#1 How to check the health status on the compute nodes and zfs heads

Using ssh and an account with superuser privileges, log into the active management node.

# ssh root@xxx.xxx.xxx.xxx
root@xxx.xxx.xxx.xxx's password:
root@ovcamn05r1 ~]#

Launch the Oracle PCA command line interface.

# pca-admin
Welcome to PCA! Release: 2.1.1
PCA>

Check the current status of the rack nodes by querying their ILOMs.

PCA> diagnose ilom

For each possible server node slot location, the Status output will show: "OK", "Not Connected", or "Faulty"
For each possible server node slot location, the Health_Details output either will show: "None" or the actual ILOM fault text.
The output is not sorted and it does not show the rack location.
In the example shown above, there is an SPX86-8003-73 fault on the node at 192.168.4.105

#2 How to identify the location of the faulted node

The output from pca-admin diagnose ilom gives the fault by node ILOM IP address.

There is no correlation between the compute node's IP address and it's rack location. The compute node IP addresses are set at provision time, and can change if and when reprovisioned.
There is a constant correlation between the node IP address and it's ILOM IP address. They are always 100 apart -- the node's IP address is always 100 less than it's ILOM IP address.
The faulted node reported at ILOM 192.168.4.105 is the same physical node at IP address 192.168.4.5.

Use the following command to grep by the node IP (not the ILOM IP) to determine the physical location.

[root@ovcamn05r1 ~]# pca-admin list compute-node | grep 192.168.4.5 | cut -f1 -d" "
ovcacn07r1

In this example, the fault SPX86-8003-73 reported at 192.168.4.105 is located in the base rack, on the compute node that is the 7th RU location up from the bottom.

Also note these two IP addresses for the zfs heads are static:
ZFS appliance ilom-ovcasn01r1 192.168.4.101
ZFS appliance ilom-ovcasn02r1 192.168.4.102

#3 How to check the health status on the two management nodes

In order to check the MN ILOMs for faults,
ssh to both MN ILOMs at 192.168.4.103 and 192.168.4.104 and type 'show faulty' to check for any faults found.

For example

[root@ovcamn05r1 ~]# ssh root@192.168.4.103
Password:
-> show faulty

-> exit

[root@ovcamn05r1 ~]# ssh root@192.168.4.104
Password:
-> show faulty

-> exit

#4 An ILOM snapshot from the faulted node may be required to assist with troubleshooting.

If needed, see How to Collect an ILOM Snapshot from a Private Cloud Appliance (PCA) or Virtual Cloud Appliance (VCA) Node (Doc ID 1591199.1)

#5 How to find additional information for any faults found.

Search MOS for the fault code.  In this example the fault code is SPX86-8003-73
Searching MOS knowledge for SPX86-8003-73 will find:
   SPX86-8003-73 - Lack of AC input power. (Doc ID 1429954.1)
MOS documents for fault codes normally will explain the following additional information:
   Type, Severity, Description, Automated Response, Impact, and Suggested Action for the System Administrator.

References

This solution has no attachment