APAR status
INTRAN
Error description
Key words: 5740XYR00 RC00E50013 MSGDSNL027I DSNLILOS . Soft cancel in the distributed job recorded in DSNL027I and DSNL028I w/ ABEND04E RC00E50013 . . From erep logrec software detail report SDWA key xDD we can find ebcan14 and ebocan14 address, for example: +012 KEY: DD LENGTH: 20 +014 00000007 93C14086 00000000 8FFFA490 |....LA +024 0F93E798 ++ 93CC8B74 ++ 93CC8B74 00000000 |.LXQL.. Use x93CC8B74 to research in MEPL for the ddf module and it normally is from DSNLILOS detecting the session loss from client deallocation. . Or you can set a slip to capture a dump and format EB (Reg 6) to get the ebcan14 and ebocan14 addresses for the issuing module. SLIP SET,ACTION=SVCD,COMP=04E,DATA=(15R,EQ,00E50013), ID=xxxx,JL=(db2subsys*),ML=5, SDATA=(ALLNUC,CSA,PSA,RGN,SQA,SWA,LSQA,GRSQ,TRT,SUM,XESDATA),END . To determine if the session is deallocated from client, we would need the following pieces of information to review. From zOS site: 1) a) TCPIP Packet trace with no wrap and sufficient allocation of the buffer. DO NOT USE ANY FILTERING ON THE PACKET TRACE. See II12014 for details on collecting a TCP/IP Packet Trace. b) Identify the client and server IP addresses (IPADDR) and PORT (PORT) numbers. c) Format the TCP/IP Packet Trace using IPCS CTRACE commands. We can request support from TCP/IP in Raleigh to assist us if we like. It may be useful to filter the raw packet trace using the known client or server IPADDR or PORT number. d) We need to find out from the formatted packet trace, where the TCP/IP FIN command originated from. To do this, search for the ACK FIN flags. If FIN command is found in the 'FROM INTERFACE' packet trace, then this indicates the client connection is the originator of the FIN command. In this case, contact the supporting team for the client connection side for further investigation. . 2) SLIP DUMP for rc00e50013 3) SYSLOG and LOGREC detail report for the timeframe From UDB client site: 1) DB2 Trace with FMT or FMTC (ddcs -c option) format file for DRDA buffer flows. 2) JCC/CLI trace depending on using the TYPE4 JDBC driver or not . The following are known situations with the rc00e50013 in DDF issued from DSNLILOS: 1) UDB DB2 CONNECT client and server via TCPIP to zOS: PMR Examples: 76286,344 92640,227 UDB DIAG.LOG recorded sql30081n communication error on Gateway: DATA #1 : SQLCA, PD_DB2_TYPE_SQLCA, 136 bytes sqlcaid : SQLCA sqlcabc: 136 sqlcode: -30081 sqlerrml: sqlerrmc: * * 0 TCP/IP SOCKETS 129.219.79.197 recv sqlerrp : SQLJCMN sqlerrd : (1) 0x81360012 (2) 0x00000012 (3) 0x00000000 (4) 0x00000000 (5) 0x00000000 (6) 0x0000000 sqlwarn : (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) sqlstate: . On the client site DIAG.LOG would have more errors. This is due to the old client udb version/fixpak while the gatew gateway is at newer UDB v81 fixpak case. UDB support suggested upgrade both client and server to current version/fixpak, ie. UDB v81 or V82 with current fixpak. Or, Client should be at least at UDB v7.2 fixpak 13 and server should be at udb v81 fixpak 8. . 2) Client is at current UDB v81 fixpak. Client diag.log recorded sql30081n with rc54 informing host (zOS) is deallocating the session. PMR Example: 80879,300,624 TCP Packet trace revealed ACK PSH FIN is from source of zOS port Normally when DSNLILOS detects the session loss, that means the client sends us 0 length data to indicate the session deallocation. In this case, there is a timing window between the cancel thread process and DSNLILOS detecting the session loss. The fix on zOS is PQ93454 / UQ93138 (db2 v71 put0410) and UQ93139 (db2 v81). . 3) JCC TYPE4 Universal Driver getting timeout resulting in RC00E50013. PMR Example: 16803,487 -> 30567,487 Client JCC TYPE4 Driver received communication error in sqlcode -4499 ( sqlcode4499 sql4499 sql4499n ). zOS DB2 server site resulted in ABEND04E RC00E50013 for soft abend issued from DSNLILOS. TCP Packet trace revealed that the ACK FIN is coming from source client port to deallocate the session. This is due to the ill defined "loginTimeout value" on the JCC03000 client. That the unexpected close of the socket occurs because JCC client has a login Timeout value that is too small. .
Local fix
Problem summary
Problem conclusion
Temporary fix
Comments
APAR Information
APAR number
II14004
Reported component name
PB LIB INFO ITE
Reported component ID
INFOPBLIB
Reported release
001
Status
INTRAN
PE
NoPE
HIPER
NoHIPER
Special Attention
NoSpecatt
Submitted date
2005-02-16
Closed date
Last modified date
2005-02-17
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
Fix information
Applicable component levels
[{"Business Unit":{"code":null,"label":null},"Product":{"code":"SG19O","label":"APARs - MVS environment"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"001","Edition":"","Line of Business":{"code":"","label":""}},{"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Product":{"code":"SSEPEK","label":"Db2 for z\/OS"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"001","Edition":"","Line of Business":{"code":"LOB10","label":"Data and AI"}}]
Document Information
Modified date:
17 February 2005