IBM WebSphere DataPower SOA Appliance unexpected reload, reboot, or an unresponsive device - Middleware News

A device can throttle for a few reasons many of which can be seen in the WebGUI Administration > Throttle settings.

This document contains three sections:

1. Questions about your appliance and environment
2. Documentation needed by IBM Support to assist in resolving the issue
3. Best practices

• Where is this device located? In other words, is it in a test or production environment?
• Are there other devices running the exact same configuration or behind the same load balancer as the device with the problem? If so how many?
• Was there any type of load test, security port scan, or other atypical traffic hitting the device at the time?
• Have there been any configuration changes, client changes, or other changes to the environment prior to the event?
• Was the appliance rebooted to recover? If so, were there any other steps taken?
• Any other recent changes related to your DataPower appliances and environment?

• Generate an error report. This can be done from the default domain Troubleshooting panel or from the command line using the following command:
  co;save error
  This will generate the error report into the device's temporary: directory for you to download and send to IBM Support.
• A full device backup is always helpful. If you have a service request (aka PMR) already open with IBM DataPower support and have recently submitted a device backup, indicate this point to the L2 support person with whom you are working.
• Any and all statistical data you have about the device leading up to and during the event. The statistics can be from SNMP, XML Management retrieval, or other methods such as command line (CLI), snapshots of the current state, etc.
• All off-device logging, again leading up to and during the time of the event. These syslog, NFS, SOAP, or other methods of off-device logging are required to help give IBM Support the proper context it requires to understand the problem.

• FFDC: If you are running firmware version 3.8.1 or later, the new device First Failure Data Capture feature is available. This feature may impact performance or throughput, but it will allow for a significant improvement in terms of isolating and determining root cause in the event of an unexpected reload or reboot. See MustGather: IBM WebSphere DataPower firmware 3.8.1 and later: Reload, Reboot, Hang, High CPU, Memory, Load, Network Failures.

• Off-device logging: This is key to debugging problems. In most cases, a specific log target per domain with the event subscription of "all" and "error" is needed. If logging is not already configured, refer to DataPower off-device logging: a configuration example.

• Statistics: Device monitoring through the command line interface (CLI) is available. This can help in a number of issues and cases from unexpected reloads, reboots, to slow response, throttling, etc. The interval at which you perform the CLI captures is very dependent on the speed at which the problem occurs. For normal operating conditions, this can be as infrequent as once every 12 hours. See MustGather: WebSphere DataPower SOA Appliance 3.7.2 and later high load, cpu, or resource use.
• Latest recommended firmware: Always review the release notes for your particular firmware version to see if there are any recently discovered issues that may relate to the issue you're encountering. See Supported firmware versions and recommended upgrade levels for WebSphere DataPower SOA Appliances.