A fix is available
APAR status
Closed as program error.
Error description
You are using the CPSM WUI or the CICS Explorer. You receive messages similar to this: . In the CPSM WUI: . EYUVC1278E Unable to obtain data for requested View. 'GET' (GET) request failed. (Failed, Exception) . or . EYUVC1278E Unable to obtain data for requested View. 'GET' (GET) request failed. (Notpermit, Usrid) . or in the CICS Explorer: . CNX0593E: RESPONSE(FAILED) REASON(EXCEPTION) . along with CNX0624E documenting a GET failure . You notice that this only occurs if the SCOPE includes a CICS region which terminated End Of Memory ( EOM ) and has not connected back to the same CMAS. . You check the CMAS EYULOG for the region which terminated EOM, and find that the CMAS does not realize that the MAS is gone yet. . In a CMAS that is having this problem, you may notice its trace filled with many entries that have debug text BADSIGNO . . Also an SVC dump of the region formatted using: . VERBX EYU9D### 'SRV' . will show the XECB field XECB_XEWK at +x'B8' is 00000000. . . Additional Symptom(s) Search Keyword(s): KIXREVSVR
Local fix
Problem summary
**************************************************************** * USERS AFFECTED: All CICSPlex SM V5R2M0, V5R3M0 and V5R4M0 * * Users. * **************************************************************** * PROBLEM DESCRIPTION: Various APARs have been opened to * * address errors that require a restart * * of the CPSM Environment Service System * * Services (ESSS) subsystem address * * space. To eliminate the need for * * customers to restart the ESSS multiple * * times, the changes to address these * * problems are being combined into a * * single PTF delivered through this APAR. * * * * The APARs addressed in this APAR, along * * with the releases they are being * * addressed in, are as follows: * * * * - PI91378 - V5R4M0 , V5R3M0 , V5R2M0 * * - PI88579 - V5R4M0 , V5R3M0 , V5R2M0 * * - PI85760 - V5R4M0 , V5R3M0 * * - PI85952 - V5R4M0 , V5R3M0 * * * * * * The problem descriptions for these * * APARs are as follows: * * * * * * PI91378 * * ******* * * When a MAS address space terminates * * with an End of Memory (EOM) failure, as * * reported by console log message * * EYUXE0028I: * * * * EYUXE0028I End of Memory entered for * * CICSPlex SM <vrm>, JOB * * <jobnm> Return Code=0000 * * * * then the CMAS it was connected to may * * not be properly notified. If this * * occurs, Toplogy Disconnect will not * * take place for the MAS, which can * * result in the following errors: * * * * - All CMASes that manage the CICSplex * * the MAS is associated with will * * treat the MAS as active and a GET * * of MAS resource table records for * * the CICSplex will show a record for * * the MAS with a status of ACTIVE. * * * * - WUI/Explorer/API/RTA requests issued * * with a Context and Scope that * * includes the MAS will fail with * * Response(1029 - Failed) and * * Reason(1359 - Exception). The CMAS * * that the MAS was connected to will * * issue the following exception traces * * for each request made: * * * * Method TPID Debug text * * CTBT 3 EXCEPT * * CTBP 13 EXCEPT * * * * - Message EYUCS0207W: * * * * EYUCS0207W Connected MAS <masnm> * * is not responding - * * APPLID(<masappl>) * * CICSplex(<plexnm>). * * * * may be issued by the CMAS the MAS * * was connected to indicating that the * * MAS is not responding. This may be * * accompanied by an RTA SAMNRM event * * being issued by the CMAS for the * * MAS. * * * * - If the MAS is restarted, it may fail * * to connect to its CMAS. When this * * occurs, the MAS will issue messages * * EYUCL0112E and EYUCL0124E: * * * * EYUCL0112E Protocol Services * * initialization unable * * to perform ICT Attach. * * * * EYUCL0124E A failure occurred * * while attempting to * * connect to CMAS * * <cmasnm>. * * * * and the CMAS will issue message * * EYUCL0107E: * * * * EYUCL0107E ESSS ICT attach to * * <masnm> failed, MAS * * already exists. * * * * indicating that the MAS already * * exists. * * * * If it does connect to its CMAS and * * if its CMAS is performing security * * checking (CMAS EYUPARM SEC(YES)), * * then all WUI/Explorer/API requests * * directed to the CMAS will fail with * * Response(1031 - Notpermit) and * * Reason(1345 - Usrid). * * * * Note that this problem is more likely * * to occur if there are multiple CMASes * * active for the CPSM release on the * * LPAR. * * * * * * PI88579 * * ******* * * - During Cache List Control Table * * (DCLT) clean-up, a bad Cache * * Alternate Index (DCAI) pointer due * * to an overlay causes an ABEND. The * * ABENDs which occur may be: * * * * - An ABEND in the EXDELGBL or * * EXDELLCL PC routine when a CMAS * * is terminated. * * * * - An ABEND in EYU0XCLY during * * execution of a CMAS or MAS, or * * during the restart of a CMAS. * * * * Additionally restart of a CMAS may * * fail with message EYUXC0010E - * * "Unable to create dataspace dspace, * * DSPSERV RC=rccccccc, REAS=rrrrrrr" * * and internal errors may occur when * * allocating cache lists that require * * a lock. * * * * - The use of an access register in * * module EYU0XCLD may cause an ABEND * * whilst attempting to issue trace * * after a failure to clear an * * Alternate Index Control Table * * (DCAI). * * * * * * PI85760 * * ******* * * ESSS startup may take longer than * * expected on systems under heavy load. * * * * The ESSS startup may exceed the 5 * * second time-out limit and if so then * * CMAS initialization will fail with * * message EYUXE0023E being issued with * * the error "Invalid return code from * * EYU9Xvrm", where "vrm" refers to the * * CICSPlex SM version, e.g. 530 * * corresponds to V5R3M0 and 540 * * corresponds to V5R4M0. * * * * In this situation, a restart of the * * CMAS usually addresses the problem. * * * * * * PI85952 * * ******* * * If a CPSM API batch job is cancelled * * while processing EXEC CPSM CONNECT, a * * S052 abend may occur in the batch job * * address space with reason code 613. * * Examination of system trace for the * * time when the abend occurred shows that * * program call ETDIS was issued at or * * near offset +247A in module EYUTXEPC. * **************************************************************** * RECOMMENDATION: After applying the PTF that resolves APAR * * PI91378, the CPSM ESSS (Environment Service * * System Services) subsystem address space * * (EYUXvrm, where for example, vrm=520 for * * CPSM V5R2M0 or vrm=530 for CPSM V5R3M0, * * etc.) must be terminated on every MVS image * * where a CMAS has executed since the last * * IPL. * * * * This can be accomplished in either of two * * ways: * * * * 1 - Insure that module EYU9Xvrm in the * * linklist concatenation is replaced by * * the updated version from SEYULINK. * * * * - IPL the MVS Image. * * * * Or: * * * * 2 - Stop all CMASes, MASes (including WUI * * servers) and CPSM batch API programs * * on an MVS image. * * * * - Use EYU9XENF to check that no address * * spaces are connected to the ESSS * * subsystem. * * * * - Use the EYU9XEUT TERMINATE function to * * stop the ESSS. * * * * - Insure that module EYU9Xvrm in the * * linklist concatenation is replaced by * * the updated version from SEYULINK. * * * * - Refresh LLA to ensure the updated * * versions of the linklist modules are * * picked up. * * * * Restart any CMASes, MASes (including WUI * * servers) and CPSM batch API programs which * * execute on the MVS image, insuring that the * * updated libraries are being picked up. When * * the first CMAS is started the EYUXvrm * * address space will be started automatically. * * * * For details on the EYU9XEUT utility, refer * * to: * * * * - CICS Transaction Server for z/OS * * - Troubleshooting and support * * - Troubleshooting CICSPlex SM * * - Tools for problem determination * * - Using the ESSS utility(EYU9XEUT) * * * * For details on the EYU9XENF utility, refer * * to: * * * * - CICS Transaction Server for z/OS * * - Troubleshooting and support * * - Troubleshooting CICSPlex SM * * - Tools for problem determination * * - Using the ESSS utility(EYU9XEUT) * * - Using the ESSS Information * * Display Utility (EYU9XENF) * * * * Note that each MVS image can be updated * * separately, and systems on an MVS image that * * are using the new code can communicate with * * systems on other MVS images that are not yet * * using the new code. * **************************************************************** The problem summaries for these APARs are as follows: PI91378 ******* When a MAS address space is terminated EOM, module EYU9XEEM, running as the CPSM EYUX subsystem EOM exit, gets control in the MASTER address space. EYU9XEEM issues a TERMINATE PC routine. During this processing, the FREE PC routine runs in the ESSS address space, to disconnect the MAS from the ESSS and to notify the CMAS that the MAS was connected to that the MAS is DETACHing. A logic error in FREE can result in the CMAS not being notified of the DETACH request. If this error occurs: - The CMAS will not issue a Topology Disconnect for the MAS, and the MAS will appear active to all CMASes that manage the CICSplex the MAS is associated with. If a WUI/Explorer/API/RTA request whose Context and Scope include the MAS is issued, the request will be delivered to the CMAS that the MAS was connected to, and the CMAS will try to deliver the request to the MAS. As part of the delivery process, the CMAS will issue a POSTECB PC routine to cross memory post the MAS. This PC routine will fail because the MAS is not active, resulting in the exception traces from EYU0CTBT and EYU0CTBP, and the requester will receive a Response of Failed(1029) and a Reason of Exception(1359). - When the MAS is restarted, an ATTACH request will be sent to the CMAS. If the CMAS cannot determine that the MAS is not already active, then the ATTACH request is failed, and the EYUCL0112E, EYUCL0124E and EYUCL0107E messages will be issued. If the CMAS can determine that the MAS is not already active, then the CMAS will issue a TERMINATE PC routine for the MAS, followed by a DETACH request. This will allow the MAS to be Topology Disconnected and Topology Connected. However, a logic error in the TERMINATE PC routine will cause all subsequent security requests issued by the CMAS to fail. As a result, all WUI/Explorer/API requests directed to the CMAS will fail and the requesters will be returned a Response of Notpermit(1031) and a Reason of Usrid(1345). PI88579 ******* - A Cache List Control Table (DCLT) in the CHE data space has a pointer which has been overlaid. This should point to the last Alternate Index Control Table (DCAI) for the cache list or to x'00000000' if the list does not have an alternate index. If the pointer is incorrect or has been overlaid, then it can result in the PC routines ABENDing and the CMAS to terminate without freeing up the global data space. The next CMAS restart may fail with message EYUXC0010E: EYUXC0010E Unable to create dataspace dspace, DSPSERV RC=rccccccc, REAS=rrrrrrrr - The use of an access register in module EYU0XCLD may cause an ABEND whilst attempting to issue trace after the failure to clear a DCAI chain. PI85760 ******* When the first CMAS of a particular release in an LPAR starts after an IPL, it creates the ESSS address space in the LPAR for that CPSM release and waits for up to 5 seconds for the ESSS initialisation code to complete before continuing. This process normally happens without a problem. However, in systems which are experiencing high system load, this 5 second wait may not be sufficient time for ESSS to complete its initialisation which may lead to the CMAS believing that ESSS is unavailable and therefore terminating its own startup. PI85952 ******* During the execution of EXEC CPSM CONNECT API command, the CONNECT PC is called to connect the caller to ESSS address space. The CONNECT PC verifies the caller's authority to connect and parameter list passed by the caller, and then create environment to call an internal BIND PC to perform the connect. Upon the successful completion of the BIND request, an ETCON is issued for all PC vectors the caller's address space is allowed to access, which connects the ETDEF tables to the caller's address space. Accordingly, during the termination of ESSS connection, an ETDIS is issued to disconnect the ETDEF tables from the caller's address space. When a CPSM API program is cancelled while connecting to ESSS address space, EOT (End Of Task) is driven in the API control subtask to release resources allocated on behalf of API program. As part of EOT processing, CPSM attempts to disconnect the ETDEF tables. If the EOT gets control before the successful completion of ETCON, abend S052 may occur in EYUTXEPC with reason code 613.
Problem conclusion
The problem conclusions for these APARs are as follows: PI91378 ******* Changes have been made to copybook EYU2XEET, which contains the code for the TERMINATE and FREE PC routines, and to copybook EYU2XERT, which contains common routines used by the PC routines, to correct the logic error that can result in a CMAS not being called for DETACH of a MAS that terminated EOM. EYU2XEET and EYU2XERT have also been updated to correct the logic error that can result in security failures when a CMAS issues a TERMINATE for a MAS. PI88579 ******* - The CPSM PC routine method EXDELGBL has been modified not to delete cache chains by following the DCAI chain. The CPSM PC routine method EXDELLCL has been modified to skip deleting the DCAI chain if an ABEND occurs due to an overlaid pointer and thus recover from such an ABEND. - Module EYU0XCLD has been modified to clear access register 1 after an internal macro call before continuing its processing. A new trace is issued to trace the EYURXCLT data structure should there be a failure in clearing the DCAI chain in module EYU0XCLD. PI85760 ******* The ESSS address space program, module EYU9Xvrm, (where vrm refers to the CICSPlex SM version, e.g. EYU9X530 corresponds to V5R3M0 and EYU9X540 corresponds to V5R4M0) has been modified to provide more meaningful messages during ESSS operations in order to assist with diagnostics should anything go wrong with ESSS startup. The following new messages are issued to the console (since the ESSS address space program does not have a joblog), by module EYU9Xvrm: EYUXE0039I <jobname> ESSS address space initialization starting. EYUXE0040I <jobname> ESSS address space has shut down. EYUXE0041I <jobname> ESSS address space starting <function>. EYUXE0042I <jobname> ESSS address space completed <function>. EYUXE0043E <jobname> ESSS address space completed <function> with an error. EYUXE0046E <jobname> ESSS module mismatch between EYU9X<vrm> (LVL <PPnnnnn> <MM/DD/YYYY>) and EYUTXEPC (LVL <PPnnnnn> <MM/DD/YYYY>). Module EYU9XECS, which is the CMAS initialization program, has been modified to wait for additional periods of time, up to a total of 1 minute for the ESSS address space to complete initialization. If the 1 minute wait time expires then EYU9XECS will issue existing message EYUXE0023E with a new insert indicating "ESSS initialization timed-out", and terminate. For example: EYUXE0023E ESSS address space not created - ESSS initialization timed-out Module EYU9XECS has also had the following new diagnostic messages added: EYUXE0038I ESSS not active - ESSS address space is being started. EYUXE0044E ESSS address space not available - CMAS terminating. EYUXE0045I RETURN / REASON / DIAG = <return> / <reason> / <diag> PI85952 ******* The following modules and copybooks have been updated: - Module EYU2XEEC has been updated to turn on flag XECB_FLAG before calling ETCON. - Module EYU2XEET has been updated to validate flag XECB_FLAG before calling ETDIS. - Module EYU2XERT has been updated to issue EYUXE0047E message and write system trace if retry is not allowed. EYUXE0047E Abend X'abend_code' in CICSplex SM version PC routine routine_name could not be retried. - Copybook EYURXEES and EYURXECB have been updated to add new flag.
Temporary fix
Comments
×**** PE18/07/17 FIX IN ERROR. SEE APAR PH00625 FOR DESCRIPTION
APAR Information
APAR number
PI91378
Reported component name
CICS TS Z/OS V5
Reported component ID
5655Y0400
Reported release
90M
Status
CLOSED PER
PE
NoPE
HIPER
NoHIPER
Special Attention
YesSpecatt / CST / Xsystem
Submitted date
2017-12-11
Closed date
2018-03-02
Last modified date
2019-05-02
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
UI54188 UI54189 UI54190
Modules/Macros
EYU0UCLD EYU0XCLD EYU9X520 EYU9X530 EYU9X540 EYU9XECS EYUTXEMS EYUTXEPC
Fix information
Fixed component name
CICS TS Z/OS V5
Fixed component ID
5655Y0400
Applicable component levels
R00M PSY UI54190
UP18/03/05 P F803
R10M PSY UI54189
UP18/03/03 P F803
R90M PSY UI54188
UP18/03/06 P F803
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":"5.2","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":"5.2","Edition":"","Line of Business":{"code":"","label":""}}]
Document Information
Modified date:
02 May 2019