Release notes for RunMyJobs releases 2024.2.0.0 through to 2024.2.0.2

Before: parameter IndependentQPJOBLOG could have 2 values true/false

After: parameter IndependentQPJOBLOG can have multiple values. true/false like the old implememtation. however also -1 (equals false), 0 (equals true) or any positive value that is the maximum file size that we retrieve QPJOBLOG.

Before: If a HTTP call didn’t get a response body back the message was always “Server sent no reply"

After: If the HTTP message is filled we show the HTTP message instead of “Server sent no reply”. Only if we do not have any message returned the “Server sent no reply”. message is printed.

Before: When using the wizard to select the procedures in packages in an Oracle database the procedures with no arguments were not visible.

After: All procedures/functions in newer oracle databases are visible in the JDBC wizard.

Before: When a PeopleSoft job is finished completely in PeopleSoft and RMJ has problems to retrieve the output the Job in RMJ is put to Error.

After: If the output retrieval, for what reason, fails it is possible to retry the retrieval of the output. This can be done by setting a ProcessServer parameter OutputRetrievalRetry or a JobParameter PeopleSoft_OutputRetrievalRetry. The parameter can have 2 options:

• {{Ignore Ignore any exception when retrieving the output, but put the job to the status according the PeopleSoft Job }}
• Retry [number of retries] (Default retry 2 times; must be 0 or greater) Retry the retrieval of output, with an optional number of retries. If the output is not retrieved within the number of retries the job is put into Error.

Before: Only a single PeopleSoft soap call could be performed at a time.

After: Multiple PeopleSoft SOAP calls are allowed.

Before: When the output retrieval fails, for what reason, a infinite loop was created to retrieve the output.

After: Depending on parameter OutputRetrievalRetry we now retrieve the output.

If the output retrieval, for what reason, fails it is possible to retry the retrieval of the output. This can be done by setting a ProcessServer parameter OutputRetrievalRetry or a JobParameter PeopleSoft_OutputRetrievalRetry. The parameter can have 2 options:

• Ignore Ignore any exception when retrieving the output, but put the job to the status according the PeopleSoft Job
• Retry [number of retries] (Default retry 2 times; must be 0 or greater) Retry the retrieval of output, with an optional number of retries. If the output is not retrieved within the number of retries the job is put into Error.

Before: In 2024.1.0.0, users without edit privilege for an object, would see a popup with an HTTP 500 error when attempting to edit that object.

After: Users see the edit page with a warning on top that they cannot edit that page.

Before: The Object Search didn’t take into account that entities may be deleted during a search, which could cause it to throw an exception.

After: The Object Search no longer has this issue.

Before: Veracode scan plan was not there in the bamboo for the RMJ Core Java.

After: It is created for the RMJ Core Java.

Before: The RMJ Eclipse-Plugin would throw a FileNotFoundException when importing a CAR file containing an xml file whose name contained an HTML encoded dot character (%2e;).

After: These CAR files can be imported and the %2e; character is decoded into the . character.

Before : System_MonitorTablesSubmitFrame and System_SynchronizeBackgroundJobsSubmitFrame was not working as expected.

After : System_MonitorTablesSubmitFrame and System_SynchronizeBackgroundJobsSubmitFrame now updated to run once per day.

Before: Earlier we were not getting correct row number in stack trace for Redwood Script, Shell Script and Library Function.

After: Now we are getting the correct row number in stack trace where we are getting the exception for Redwood Script, Shell Script and Library Function.

Current Behavior: No ad-hoc alert was being raised when operator message for the job was already served.

New Behavior: Now an ad-hoc alert will be sent per ad-hoc alert source for the job.

Before: The default value of DatabaseRetries configuration was 18 and the max value was 45.

After: Both values have been updated. Now the default value is 100 and the max value is 1000.

Before: It is possible due to timing that a step is going into Error if the job cahin calls has a OnChange Trigger/Action attached.

After: This has been resolved.

Before: the upgrade logic for deduplicating process definition parameter display order could slow down the upgrade process significantly.

After: this code has been optimised.

Before: A number of three-letter time zones(e.g. EST, HST, MST) have been deprecated by Java a number of years ago. An upgrade to a recent version of RMJ may cause usages of these time zones to be replaced by null values. In particular in the context of recurring processes. These null values could lead to NullPointerExceptions during runs of the System_ProcessKeepClauses maintenance process.

After: Every JobGroup’s timezone with a null value will be updated from Job’s timezone if it is not null in table. Otherwise it will be set to GMT. A new operator message for all the JobGroup will be shown to user about JobGroup’s timezone updation.

Before: Selecting multiple objects (Job definition, Job chain, Queue, etc.) and exporting them is no longer working.

After: This issue should be resolved.

Before: Description of standard PeopleSoft job definitions was confusing.

After:

System_PeopleSoft_RunJob

Before:

This is a template to run PeopleSoft Process Definitions. This template can be used to run PeopleSoft Process Definitions already defined in the PeopleSoft system.

After:

• description: Process Definition for running PeopleSoft jobs
• comment: You can use this Process Definition to run jobs defined in the PeopleSoft system. Only the PSJob job type is supported.

System_PeopleSoft_RunProcess

Before:

This is a template to run PeopleSoft Process Definitions. This template can be used to run PeopleSoft Process Definitions already defined in the PeopleSoft system.

After:

• description: Process Definition for running PeopleSoft processes
• comment: You can use this Process Definition to run PeopleSoft processes. All process types other than PSJob are supported.

Before: Latest user was not necessarily the one that was re-enabled when using the UserAdmin API.

Before: Latest user is the one re-enabled when using the UserAdmin API.

Before: Arithmetic and logical operators cause an error in an array REL parameter.

After: It should allow to saving of array parameters with arithmetic and logical operators.

Before: Veracode scan plan was not throwing exception for the RMJ Core Java.

After: It should not throw any exception.

Before: the DB2 JDBC driver that was shipped with the product contained the following vulnerabilities: CVE-2023-27859, CVE-2023-45193, CVE-2023-47141, CVE-2023-47145, CVE-2023-47152, CVE-2023-47158, CVE-2023-47746, CVE-2023-47747, CVE-2023-50308

After: the driver has been upgraded from version 11.5.7.0 to 11.5.9.0, which resolves these issues.

Before: BOBJ process server could shutdown if CMS was temporarily unavailable

After: BOBJ process server will not shutdown if CMS was temporarily unavailable

Before: FL_FILENAME job parameter in duplicated SAP_CcmsJobGetSpoollist has been ignored

After: In value of FL_FILENAME job parameter in duplicated SAP_CcmsJobGetSpoollist is considered, in value is now used as prefix for spool file names and out value (if the parameter is set to in/out) is filled with job files references

Before:
When System_Upgrade job is executed and no catalog-admin is present in catalog-service, then the job ends up in “Error” status.

After
When System_Upgrade job is executed and no catalog-admin is present in catalog-service, then the job ends up in “Completed” status. stdout.log includes message Catalog service returned an error (HTTP 404).

Before: Job description translations of duplicated CcmsJobGetoutput job definitions are missing

After: Job description translations of duplicated CcmsJobGetoutput job definitions are corrected

Before: RMJ was running on Tomcat 9.0.83, which is vulnerable to CVE-2024-23672 and CVE-2024-24549.

After: RMJ is now hosted on Tomcat 9.0.87, which is the latest release and without known vulnerabilities.

NOTE: The configuration of RunMyJobs limits the impact of these vulnerabilities, making the believed risk of exploitation low, however the version is being upgraded out of precaution.

Before:
jgetfile uses a fixed connection timeout of 10 minutes.
After:
Commandline parameter -timeout <seconds> is added to jgetfile to enable setting the connection read/write timeout. The default value is still set at 600 seconds (10 minutes).

Before:
File events on Linux Platform Agents could be missed if the file creation timestamp was in the future.
After:
 File events work correctly, including the case when a file contains a future timestamp.

Before: Vulnerability scanners sending excessively large chunk transfer encoding or excessively large Content-Length HTTP requests to a Platform Agent would result in ambiguous StringBuffer maximum capacity exceeded FATAL messages being logged.

After: Vulnerability scanners sending excessively large HTTP requests to a Platform Agent now result in clear ERROR messages being logged.

Before: Platform Agent was built with OpenSSL version 3.0.9
After: Platform Agent is built with OpenSSL version 3.0.13

Before:
Startup of Platform Agent on Unix systems, could hang if there are no older version directories present.
The log file would repeatedly show the following line in debug mode:
DEBUG 2024-02-23 13:29:47,318 CET 12044-sendmessage-9 #3 opsys.dir - Creating directory /opt/redwood/agent/.lock

After:
Platform agents starts up normally when there are no older versions present to be cleaned up.

Before: After importing a multi-step SAP jobs as single definition, this imported SAP job can only be submitted with all the steps

After: Import now adds a new parameter DISABLE_STEPS which allows selection of steps which should be disabled before submission, ie. these steps will not be created on the submitted SAP job.

Before: Big size PDF spoollist cannot be retrieved properly due to timing issue with spool host.

After: Timing issue has been fixed.

Before: There are several issues by retrieving spool file with spool host and SAP_CcmsJobGetSpoolList:

1. spool file is retrieved with connector not spool host
2. big pdf spool file cannot be retrieved.

After: These issues have been fixed.

Before: Job completed successfully even if RTX parsing failed

After: Set job to error with notification if RTX parsing failed

Before: the PostgreSQL JDBC Driver (version 42.5.3) that was shipped with RMJ contained CVE-2024-1597.

After: the PostgreSQL JDBC Driver was updated to version 42.7.2, which contains a fix.

Before: When you create a step in the chain and add Locks, Raise, and Wait events. And move this step to another position your locks and events will disappear.

After: Now when you create a step in the chain, add Locks, Raise, and Wait events. And move this step to another position your locks and events will still be there

Installation that made use of custom home pages (f.e. FCA) it was required to have the scheudler-it-role to be able to access the control center. This requirement has been removed, and everyone can now access the control center. If this is not wanted, the system can be configured to only allow users that have the Access_Control_center global privilege to access the control center.

Before:
The interface showed an error in case that no "/configuration/jcs/ui/helpurl" registry entry was defined. This impacted the “Preview new UI button”.

After:
Default registry value will be used and no error occurs when previewing the new UI.

Before: Showing the details page of a Job Definition with a huge source could potentially cause the browser to crash.

After: The Source of a Job Definition shown on its detail page is truncated to 100,000 characters, preventing the browser running out of memory. The full Source is still visible on the edit page of the Job Definition. As a side effect this makes these detail pages more performant, both when opening and when performing actions.

Before: The project ‘redwood-scheduler-doc-manuals’ was build using ant and markdown files where generated for html documents.

After: The project ‘redwood-scheduler-doc-manuals’ now build using Gradle and now it will directly produce html without using ant.

Before: A Job Datum is always copied when a Job is copied.

After: The Datum Definition can specify with CopiedToNewJob if Job Data should be copied. The default is to copy it.

Before: Operator messages are created too often when spool host can't connect to the SAP system

After: Repeated operator messages are created only every 10 minutes when spool host can't connect to the SAP system

Before: All users had access to the RunMyJobs control center.

After: A new global privilege, Access_Control_Center, has been added. By default this privilege is not activated. When this is activated, then a user must have this global privilege assigned to them in order to be able to access the control center.

Only users with assigned global privilege Access_Control_Center will have access to the RunMyJobs control center. This global privilege is assigned to the scheduler-administrator and redwood-administrator roles by default.

NOTE: This can only be activated by an administrator, and then only via scripting. In order to activate this, execute the following snippet in the shell:

{
  GlobalPrivilege privilege = jcsSession.getGlobalPrivilegeByName("Access_Control_Center");
  privilege.setActivated(true);
  jcsSession.persist();
}

Before: File events on Linux could be missed, (most likely) when the creation timestamp of these files were in the future.
This caused that such files would not be picked up by the notification system anymore when the file had ‘aged’.

After: Files events are picked up and handled as before, using the polling mechanism.