Service and maintenance

Edit online

Many secure accounts are hesitant about enabling remote service. Consideration should be given to enabling outbound RSF calls that contain the data necessary to automatically dispatch a service representative. Since there is considerable customizing capability provided, RSF can probably be tailored to match your installation's security policy and practices.

This product has support for the concurrent service and maintenance of hardware. The following can be serviced concurrently while normal customer operations continue:

  • Power supplies
  • Channel cards
  • Licensed Internal Code (LIC)
  • Processor Drawer

When service is performed on the above-listed elements of the processor, the physical elements are logically and electrically isolated from the remaining portions of the system still in use. This is begun by first logging on the HMC with a SERVICE ID and then performing the desired maintenance or service task. Refer to the Service Guide for information on how to perform the desired task.

Note: Before placing a reconfigurable or shared channel path into a service configuration, record the logical partition name(s) that it's currently assigned to. This will assure that after service is complete, the channel path will be returned to the logical partition(s) to which it was previously allocated, even if different operations personnel are now in charge.

When a partial or complete configuration is surrendered for service or maintenance the following recommendations should be followed:

  • The IOCDSs should remain write-protected.
  • All installation configuration data should be, or has been previously, saved. The installation configuration data should be restored, and the initial activation must be fully manual. When activation completes, use the following procedure to check the active I/O configuration:
    • go to the Configuration task list
    • select Input/Output (I/O) Configuration task
    • select the IOCDS which is marked Active
    • select the View pulldown
    • in the View pulldown, select Channel Path Configuration
  • Prior to giving the system to the service representative for disruptive maintenance, it is advisable to idle the partitions (perform an orderly shutdown of the applications and control programs occupying the partitions, followed by stopping each partition) rather than deactivating them. Doing this allows the system to perform automatic (re)activation on the subsequent activation. Automatic activation offers fewer opportunities for human error to affect the controlling parameters of the system, and hence is more secure.

After completion of a disruptive service operation, the CPC should be selected and the I/O Configuration task invoked to display a list of the available IOCDSs. Use this list to check the IOCDS time stamps against the values recorded the last time the IOCDSs were updated. This is to help ensure that the IOCDSs remain unchanged.

The Hipersocket Sniffer debugging tool should not be used in normal operations, as it could cause security problems. For more information, see the IBM® HiperSockets Network Traffic Analyzer (HS NTA) Frequently Asked Questions document, found at https://www.ibm.com/search?lang=en&cc=us&q=hipersockets.