Diagnosing an excessive spin (spin loop)

To avoid unnecessary recovery actions, system functions that can validly exceed the interval are exempt from excessive spin-loop processing, so that they will not cause any recovery actions. If they exceed the time-out interval, these system functions do cause an excessive spin loop record to be written to the logrec data set.

You can examine the in-storage logrec buffer for entries that recovery routines have made, but which were not written to the logrec data set because of a system problem. Very often it is these records that are the key to the problem solution. See the topic on obtaining information from the logrec in z/OS MVS Diagnosis: Tools and Service Aids.

When the system writes a dump, the dump includes the records in the logrec buffer in storage; the buffer records have been either written to the logrec data set or are queued to be written to the logrec data set. When you begin to diagnose a dump for a system problem, you can use IPCS to view the system records in the logrec recording control buffer.

Before you begin You need to have IPCS installed and the following information:

An EREP report of SYS1.LOGREC
Any SVC dumps, stand-alone dumps showing the ABEND071, or SYSMDUMP dump

SYSLOG and OPERLOG. Look for spin loop messages like IEE331A or IEE178I. Note that IEE331A is rare; IEE178I is more likely to occur. For example:

00 IEE331A PROCESSOR (0) IS IN AN EXCESSIVE 
DISABLED SPIN LOOP WAITING FOR CPU IN 
STOPPED STATE 
REPLY U OR SPIN TO CONTINUE SPIN 
REPLY ABEND TO TERMINATE WORK ON 
PROCESSOR (1) WITH RETRY, 
REPLY TERM TO TERMINATE WORK ON 
PROCESSOR (1) WITHOUT RETRY,
OR STOP PROCESSOR (1) AND 
REPLY ACR 
(AFTER STOPPING THE PROCESSOR, DO NOT START IT)

IEE178I AUTOMATIC RECOVERY IS IN PROGRESS NO OPERATOR ACTION IS REQUIRED 
PROCESSOR (00) DETECTED AN EXCESSIVE DISABLED SPIN LOOP
WAITING FOR LOCK RELEASE FROM PROCESSOR (03). 
AUTOMATIC RECOVERY ACTION IS SPIN