A fix is available
APAR status
Closed as program error.
Error description
CPSM uses message buckets to communicate certain information between regions (CMAS-to-CMAS or CMAS-to-MAS.) If CPSM goes to retrieve an empty bucket and none are available, it will write an exception trace entry and keep a count of how many times this 'No Buckets' situation occurs during the region run. This trace entry though may be overwritten if the trace wraps. This APAR is being opened to add additional external messages when the XDATLOST condition occurs, so that it can be made available for problem determination if needed. . Additional Symptom(s) Search Keyword(s): KIXREVGJT
Local fix
Problem summary
**************************************************************** * USERS AFFECTED: All CICSPlex SM V4R2M0 Users * **************************************************************** * PROBLEM DESCRIPTION: Your CICSplex occasionally fails to * * capture CICS notifications, which * * in turn give rise to other issues * * within the CICSplex. Such errors * * may include (but are not limited to): * * * * - failure to recognise instances of * * newly defined CICS resources, * * - failure to recognise an Operator * * Sign-off, causing WLM affinities to * * persist for too long. * * * * These conditions can occur without any * * visible warning to the normal message * * logs, and are only detectable if a * * CICS auxiliary trace is available * * that has not wrapped back over itself. * * * **************************************************************** * RECOMMENDATION: After applying the PTF that resolves * * this APAR, all MASes and WUI regions * * must be recycled to pick up the new * * code. Note that regions do not need * * to be brought down and restarted at * * the same time. * * * **************************************************************** Notifications of events from CICS to CPSM are captured via CPSM provided GLUE modules. These modules capture CICS data into predefined data "buckets", which are swapped from a free chain to an in-use chain. The elements in the in-use chain are gathered and emptied by CPSM long running task processes, which then make the appropriate adjustments to the CPSM operating environment commensurate with the original CICS notification. The "used" bucket is then swapped from the in-use chain to the free chain, ready for subsequent re-use. On occasions of intense CICS notification activity, for instance - if a large number of CICS resources are being installed or discarded simultaneously - then it is possible for the free bucket chain to be emptied, meaning some notifications are lost. CPSM registers this occurrence by taking special exception traces, but these will be lost as soon as the auxiliary trace dataset wraps around, leaving no clue as to why CPSM is possibly behaving erroneously. In addition, if a lost bucket involves the notification of a new (or updated) CICS connection, then the GLUE module will set a flag to inform the CPSM MAS heartbeat process to initiate a scan of all SNA and IP connections defined to the CICS region. This should cause CPSM's Topology data for the region to be consistent with what is actually installed. If the SNA connection scan should fail unexpectedly, it can cause the IP connection scan to be skipped. In turn, this will prevent consistency between CPSM Topology data and the actual IP connections active in the MAS. One spin off effect of this will be that Workload Manager may choose an inappropriate dynamic routing target region, when an acceptable IP connected target is available.
Problem conclusion
Changes were made to MAS exit data processor (EYU0NLXD) to issue a new CPSM message on the first occasion that a "no buckets" condition arises: EYUNL0144W -applid- MAS notifications have been lost - no free message buckets. The message explanation will advise customers that see this message to extract and print the auxiliary trace datasets of the reporting region as soon as possible, and consider increasing the value of the MSGBUCKETS EYUPARM in the MAS if the condition persists. The CPSM XRSINDI Exit module (EYU9NLID) and the XMEOUT Exit module (EYU9NLME) have been modified to employ separate flags for notifying of SNA and IP connection notification failures. Correspondingly, the Topology Resource synchronizer (EYU0NLT1) has been modified to use both of the flags to control the scanning of installed SNA and IP connections in a MAS.
Temporary fix
FIX AVAILABLE BY PTF ONLY
Comments
APAR Information
APAR number
PI54396
Reported component name
CICS TS Z/OS V4
Reported component ID
5655S9700
Reported release
70M
Status
CLOSED PER
PE
NoPE
HIPER
NoHIPER
Special Attention
NoSpecatt / Xsystem
Submitted date
2015-12-17
Closed date
2016-03-29
Last modified date
2016-04-05
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
PI54986 UI36561
Modules/Macros
CJA9NLID CJA9NLME CJB9NLID CJB9NLME CJC9NLID CJC9NLME EYUMCNLC EYUMCNLE EYUMCNLK EYU0NLT1 EYU0NLXD EYU9NLID EYU9NLME
Fix information
Fixed component name
CICS TS Z/OS V4
Fixed component ID
5655S9700
Applicable component levels
R70M PSY UI36561
UP16/03/31 P F603
Fix is available
Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.
[{"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SSGMGV","label":"CICS Transaction Server"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"4.2","Edition":"","Line of Business":{"code":"LOB35","label":"Mainframe SW"}},{"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Product":{"code":"SG19M","label":"APARs - z\/OS environment"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"4.2","Edition":"","Line of Business":{"code":"","label":""}}]
Document Information
Modified date:
05 April 2016