JOBWATCH Release Notes

v1.00  - Original

v1.01  - Fixed bug when using *ALL for subsystem and debug mode.  Detail 

          lines weren't printing.

         Fixed heading alignment on debug report.

v2.00  - Registration updates.

v2.01  - Registration fix.  If you aquired keys for v2.00 of JOBWATCH, please contact 

          us for new keys for this version.

v3.00  - Added Job Maintenance screen (F6 from JOBWATCHCF command) that allows you to

          set up jobs to check for activity.  Enter a job name, a status you want to

          check for (*ACTIVE, *JOBQ, *OUTQ, or *ANY for every status) and the minimum

          number of jobs matching that status to check for.  If less jobs that 

          specified are found, the command entered will be run.  

v3.01  - Parameters &Y and &L were reversed.  Fixed.

         Added new parameter &E that will return then message text from the last (newest)

          message from the job log in question.

        (SP111006) - Updated message returned from &E to include the message

          help/cause.

         Fixed bug in message returned with garbage characters.

         More error checking in returning error message text.  Was

          erroring when it tried to process a job that no longer

          existed.

         Fixed bug when returning an error message greater than

          1024.  Was causing an error in F.MSG that said length

          or start position was greater than the receiver variable.

         Fixed bug when using &E code and message had a single quote

     in it.

    Added feature in F6 (Job Maintenance) so in addition to 

          specific job names to look for, also allow the use of

          wildcard values. 

v3.10  - Added new option in the Job Maintenance screen to allow the

          status of *MINACT.  This will stand for "Minutes Active".  The

 MIN column can then be filled with the maximum number of minutes

 you want a job to run.  If this is exceeded, the command 

 associated will be run.

v4.00  - Added licensing by date (SaaS) option.

v4.10  - Added the replacement variable &C which will return a list of 

            object locks for the job in the format:

            (library,object,type,status)(library,object,type,status)....

           Removed &K replacement variable.

           Included module JW001MD that has a procedure named JWObjLocks()

            that will return a list of object locks for a specific Job,

            user and job number.

           Included sample program JWSAMP that shows how to use the

            JWObjLocks() procedure.  

           Included source PF QRPGLESRC that has source for JWSAMP

           Included source PF QCOPYSRC that has the /COPY member for

            JW001MD prototype definition(s).

           Made replacements bigger for message so that the message isn't

            truncated.   

           SP170405 - Bug found when a job had a lock on more than 100 objects. Fixed.

           SP170421 - Added the following replacement parameters for the *MINACT 

            Job Maintenance commands.  &U - User ID, &N - Job Number, &T - Job Type, 

            &B - Job Subtype

v5.00 -   Added Message Queue Status monitoring (V7R2 or higher required).  When using

           JOBWATCHCF the F7 command key will allow you to set up Message Queue 

           monitoring by message queue, user,job, severity, etc.  The replacement 

           variables avilable are:

            &L = Message Queue Library (10)

            &Q = Message Queue (10)

            &I = Message ID (7)  

            &T = Message Type (13)

            &S = Message Subtype(22)

            &X = Message Text (1024)

            &V = Message Severity (5)

            &Z = Message Timestamp (26)

            &U = From User (10)

            &J = From Job (28)

            &P = From Program (10)

            &R = Message File Library (10)

            &F = Message File (10)

            &2 = Message Second Level Text (4096)

            &M - Machine Name (8)

          Added a TRIM option for commands that are run.  By default TRIM will be

           set to *YES which means the values used in substitution will be trimmed.

           This option is good when you are calling a command.  The value of *NO

           will not trim the data.  This option is good when you are calling a  

           program with replacement variables.

          Changed the debug option to produce a stream file in the IFS vs a printer

           file.  The default file name is /tmp/jobwatchdebug.txt and can be

           changed on the JOBWATCH command.           

          SP200821 - Bug found when retrieving job log messages.  For one call the

            Clear API wasn't being called after which could have caused

            storage issues in some cases.

          SP210412 - Updated application to allow 1000 watches vs 40 previously.

          SP210422 - Updated debug log to show number of jobs running vs number

            of jobs processed.

          SP210505 - Added more debugging messaging for subsystems monitored, type

            of jobs and *ACTIVE and *MINACT jobs.

          SP210507 - Added a check when *ACTIVE match is met to double check the 

            total number of jobs running.

v5.10 -   Removed &U, &N, &T, &B and &I from being used as replacement variables

           when checking for a ACTIVE job counts.  That's because these 

           really aren't available to the system when counting the number of 

           jobs.

          Added option in Job Status Maintenance to allow the monitoring 

           of CPU percentage.  For example, no monitor for jobs over 50% CPU,

           you would use the entry *C50.  The command run will then be done

           when a job's CPU% is greater than or equal to 50.

          Added &CPU as a replacement variable for CPU and Status jobs.  The

           job's CPU percentage will be replaced (9,2).

          When running multiple cycles, every 20 cycles the job CPU statistics

           will be reset so that the job list being used to calculate the CPU

           percentage is cleared of any ended jobs.

          Added a message to alert the highest CPU job at the end of a run in        

           the JOBWATCH debug output.

          SP210617 - Fixed bug when number of jobs was over 9,999.   

v5.20 - Added Monitor Message Queue (MSGQ) parameter to JOBWATCH command.  This

         defaults to *YES.  This can be useful when submitting multiple

         JOBWATCH jobs and you want some to not monitor a message queue.

v5.30 - Updated to include option *ONLY for the MSGQ parameter so that you 

         can submit a job that only watches message queues.

         SP241220 - Updated job lock procedure to use new SQL feature in V7R3 vs

          API.

         Added more logging to debug file for object locks.