Troubleshooting
Problem
This document provides a good introductory description of the Work Control Block Tables.
Resolving The Problem
The work control block table (WCBT) is also called the Job Table. This document provides a general description of the WCBT and includes a variety of additional items that should be helpful in dealing with and understanding problems and issues relating to the WCBT. Having this information collected together should help make it more readily accessible.
|
o | The WCBTs are tables of entries, each of which provides part of a permanent nucleus in the system for each job that is active, on a job queue, or has spooled output. |
| o | The WCBTs are created during the IPL when they do not exist or are damaged. A cold start IPL (a clear/clear/clear on the IPL options screen) will result in the WCBT being deleted. This will allow users to get the storage back if the table becomes too large for their system. |
| o | If any of the WCBTs become damaged, they are all deleted. The objects that must be deleted when the tables are deleted are deleted later in the QWCBTCLNUP job rather than during the IPL path. |
| o | The WCBTs are permanent extendable space objects. Each WCBT can hold 16352 job structures. |
| o | The addresses of the WCBT objects are stored in the QSYS library. |
| o | The WCBT header object is named QWCBT00. The WCBT objects that contain the nucleus job structures are named QWCBT01 to QWCBT10 up through the R450 release. This was extended for the R510 release to handle tables QWCBT11 through QWCBT30. The number of WCBT objects that can actually be created for a system is a function of system value QMAXJOB which was added for R510 release. This system value is shipped set to 163520 which reflects the maximum job structures for 10 WCBTs. Therefore, the maximum number of jobs on systems up through R450 is 163520. The upper limit on QMAXJOB is 485000. This limit is somewhat smaller than what 30 tables can hold. The reason for the value being somewhat smaller than 30 x 16352 or 490560 is that the system allows a buffer so that the system is able to IPL even when the maximum limit is reached. |
| o | The WCBT is extended when all previously created WCBT job entries are assigned to a job (in use). The number of WCBT entries added to the system when it is extended is controlled by the QADLTOTJ system value. |
| o | The DSPJOBTBL command shows information about the number of entries in the WCBT. This command was first made available at release R410. |
| o | The WCBT structures can be viewed using display/alter storage through the system service tools. The object type is Space (19) and subtype is D0. Context is QSYS. |
| o | The permanent system jobs will always be in the first entries in QWCBT01. The IPL code always places SCPF in the first job structure and the other permanent system jobs in succeeding structures until they have all been created. If joblogs are needed for the system jobs from the last IPL cycle, the IPL code will relocate the job structures for those prior instances of the system jobs so that the job logs and other spooled files will not be lost. |
| o | These permanent job structures will be made available for reuse after their jobs have ended and their spooled files have all been deleted. |
| o | A WCBT compression operation will compress the various WCBTs such that all job structures for jobs still on the system will be moved down to the bottom of the tables and extra entries will be removed from the top. For example, if there are 4 tables full of entries but there were only existing jobs for less than two full tables, a compression would result in the existing job structure entries to all be moved down to the first two tables and the latter two tables would be deleted. The Rochester Support Center knowledgebase document New, Slow IPLs/Work Control Block Table - Compression: describes how these compression operations are accomplished. |
| o | CPI1468 messages are sent when the number of jobs in the system is more than 90% of the configured maximum (163520 for releases up through R450) number of jobs allowed. Note: Rochester Support Center knowledgebase document N1017161, CPI1468 Messages - When and Where Sent describes this in more detail. To link to document N1017161 immediately, click here  . |
| o | An IPL of the system can fail with an SRCB9003610, B9003610, if it determines that the system is out of job table entries. An IPL can also hang indefinitely without failing at SRCC9002A90, C9002A90, (Starting system jobs) if the SCPF code is waiting on creating of additional job structures to use for starting the system jobs. A clear/clear/clear manual mode IPL can be used to successfully start the system following such a failure or hang. In current releases spoolfiles will not be deleted but will be detatched from the job. |
[{"Type":"MASTER","Line of Business":{"code":"LOB57","label":"Power"},"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SWG60","label":"IBM i"},"Platform":[{"code":"PF012","label":"IBM i"}],"Version":"7.1.0"}]
Historical Number
25290605
Was this topic helpful?
Document Information
Modified date:
18 December 2019
UID
nas8N1019426