Signaling sympathy sickness

signaling sympathy sickness can occur to a sending system when one or more members become stalled on the target system. The stalled target member fails to recycle its inbound buffers and therefore cannot receive signals properly. If signals are not being received, outbound signal buffers on the sending system remain in use. If the situation persists, XCF on the sending system will eventually run out of the outbound buffers and any further IXCMSGO requests will be rejected.

When sympathy sickness is detected, message IXC631I is issued by the system causing the problem to indicate that a local stalled XCF member appears to be causing sympathy sickness on some other system in the sysplex. This message is displayed once for each system impacted by the stalled member.

The system causing the sympathy sickness will also issue one instance of message IXC640E to indicate that a problem exists and to describe a method of resolution. However, both message IXC631I and message IXC640E may not be displayed if consoles on the sick system has been impacted and never processed the messages.

Message IXC440E is issued by the impacted system to point out which system was causing the sympathy sickness. Note that even if message IXC640E is not displayed on the faulty system, message IXC440E can still be displayed on the impacted system. Thus it is recommended that operational procedures or system automation be updated to deal with message IXC440E. See z/OS MVS System Messages, Vol 10 (IXC-IZP) for more information on these messages.

You can use the DISPLAY XCF,GROUP command to investigate stalled members and signaling sympathy sickness if manual intervention is required. You can also use the MEMSTALLTIME parameter in the SFM policy to allow SFM to take automatic action to alleviate signaling sympathy sickness. See Handling signaling sympathy sickness for more information about defining an SFM policy for sympathy sickness.