Hot I/O detection

Hot I/O refers to the repeated I/O interruptions that result from hardware malfunctions. Because it can cause the system to loop or to fill the system queue area with I/O control blocks, hot I/O needs to be detected quickly and corrected.

When the number of repeated interruptions exceeds an installation-defined threshold value, the system assumes there is a hot I/O condition. If your installation has set up hot I/O recovery defaults that the system can use, the system issues message IOS109E and attempts to recover from the hot I/O condition. (See the IECIOSxx parmlib member in z/OS MVS Initialization and Tuning Reference.) If your installation has not set up hot I/O recovery defaults, the system issues one of the following messages, if possible, or loads one of the following restartable wait states and prompts you to take action:

IOS118A or IOS111D  —  HOT NON-RESERVED DIRECT ACCESS DEVICE
(Wait state 111)
IOS119A or IOS112D  —  HOT RESERVED DIRECT ACCESS DEVICE
(Wait state 112)
IOS117A or IOS110D  —  HOT NON-DIRECT ACCESS DEVICE
(Wait state 110)

When you take action, try to solve the problem at the lowest possible level. That is, try to correct the problem at the device first and then at the control unit. You could power the device off and on. If that does not help, you could reset the control unit if the affected device is not a direct access device. If these actions do not correct the problem, you might have to physically disconnect the device or control unit.

Whatever action you take, tell the system what you are doing by responding to the prompting message or restartable wait state. See z/OS MVS System Messages, Vol 9 (IGF-IWM) for information about IOS messages, and z/OS MVS System Codes for a detailed explanation of the restartable wait states and your response to them.

Hot I/O describes how z/OS® handles a hot I/O condition.