Manager Policy Group
The Manager policy group contains the following software delivery manager policies. You can modify policy parameter values by double-clicking a policy to display the Setting Properties dialog.
Configures the movement of installation records. There is a possibility of records being present on two domain managers from moves prior to an upgrade to Unicenter Software Delivery 4.0, when move functionality was not present in the product. In these cases, you have the choice of the following move methods:
Installation records from the previous domain manager are merged with any existing records on the new domain manager.
Installation records found on the new domain manager are retained, and those found on the previous domain manager are disregarded.
Installation records found on the new domain manager are replaced with those found on the previous domain manager.
Default: 2
Controls the move functionality of a domain manager when one or more of its downstream computers execute a move operation, initiated by having their server address changed. If set to False, move functionality is disabled and moves are not even attempted. If set to True, move functionality is enabled and the domain manager always attempts to move installation records from the previous domain manager.
Default: True
Controls the handling of any ongoing jobs that are found on the previous domain manager when a computer move is to be initiated. If set to False, any ongoing jobs found on the previous domain manager at the time of the computer move are not moved. Instead, another attempt will be made the next time the computer mover task is scheduled. If set to True, all completed job records are moved, ignoring the fact that ongoing jobs may exist on the previous domain manager. The records for ongoing options are destroyed.
Default: False
Specifies whether to move the OSIM job information along with the installation records from a previous domain manager during an SD computer move operation. You can set the parameter value to True or False. Setting the parameter value to True always attempts to move the OSIM job information and the installation records from the previous domain manager. However, setting the value to False disables transfer of the OSIM job information, and attempts to move only the installation records.
Default: True
Specifies whether to move the installation records from a previous domain manager after initiating transfer of the OSIM job information. If the transfer of the OSIM job information fails with an error code that is present in the error codes list specified for this setting, the transfer of the OSIM job information and the installation records is postponed until the next computer move attempt is made. However, if the transfer of the OSIM job information succeeds or fails with an error code not included in the list specified for this setting, the process of moving the installation records is initiated.
Note: The error codes list is a comma-separated list and can include the error codes: 1,2,3,4,5,6,8,9,12,13,45,107,207,211,214. For more information on these error codes, see the Error Codes section in the Software Delivery help.
Default: 1,2,3,4,5,6,8,12,13,45,211
Specifies the total time in minutes that the computer mover task runs before giving up the processor to other tasks. Valid values are non-negative integers.
Default: 10
Specifies the time in minutes between two runs of the computer mover task. This value is always overridden when a new move is scheduled. Valid values are non-negative integers.
Default: 30
Specifies the CA MDB database timeout period in seconds. A value less than the default of 300 (5 minutes) is permissible, but larger values are not permissible.
Default: 300
Specifies the maximum number of minutes an API server is permitted to be idle before being shut down automatically.
Default: 360
Specifies the maximum number of seconds a session can be idle in bound mode before the API server forces the session to be unbound. The API client will be notified that the API server is unbound and will automatically re-bind to an API server upon the next API request.
Limits: 10–120
Default: 120
Specifies the maximum number of API servers in the API pool.
Default: 10
Specifies the minimum number of API servers in the API pool.
Default: 1
Specifies the maximum number of threads for handling SDCatalog API requests.
Default: 50
Specifies the number of threads for handling ordinary API requests.
Default: 10
Specifies the time in hours before a container cleanup job starts. After distribution, a container remains in the list displayed in the Main - Distribution Jobs window for 168 hours (7days) by default.
Note: The analysis is done every four hours; do not specify a value less than 4.
Default: 168
Specifies the time in hours that an item that is fetched from an area remains in the list in an Area-Fetched Items window. The default is 336 hours (14 days).
Note: The analysis is done every four hours; do not specify a value less than 4.
Default: 336
Defines the maximum time in minutes to send distribution acknowledgement files to the enterprise manager from a domain manager. This time definition specifies how long the Task Manager attempts to send the acknowledgement file to the enterprise manager, counted from the point when the corresponding container was received, unpacked, and all orders executed.
Default: 120
Indicates the maximum time in minutes for sending container files from the enterprise manager to domain managers. This time period specifies how long the Task Manager attempts to send the file to the domains, starting from the time when the corresponding container was activated, packed, and ready for shipment. After this period of time, the areas are set to an error state.
Default: 120
Specifies the maximum distribution time in minutes for sending distribution files to domain managers from the enterprise manager. This time period specifies how long the Task Manager attempts to send this file to the domains.
Default: 360
Specifies the number of minutes to wait for a DTS file transfer to start. If the transfer does not start within this interval, the DTS file transfer is abandoned. If this value is set to zero (0), the wait time is unlimited. Valid values are all positive integers.
Default: 0
Indicates whether license checking is skipped. If True, no check for valid licenses is performed; instead forced restrictions on non-CA packages are applied. If False, license checking is carried out. Note that this will initiate the 30-day grace period if no valid Software Delivery licenses are installed. You may use this 30-day period to trial the full Software Delivery functionality.
Default: <locally managed>
Indicates whether the reboot event message should be skipped each time an SXP installation initiates a system reboot. If True, this particular event message is turned off. If False, an event message is sent whenever the installation is interrupted by a reboot request.
Default: False
Controls the frequency at which you want to perform the cache cleanup for the Installation Manager.
Default: 5
Limits: 0 to 10080
Controls the frequency at which you want to perform the cache cleanup for the Notification Manager.
Default: 60
Limits: 0 to 10080
Controls the frequency at which you want to perform the cache cleanup for the Policy Manager.
Default: 60
Limits: 0 to 10080
Specifies the maximum number of high priority messages from servers to be processed in an sdmgr_im processing cycle.
Default: 2
Specifies the maximum number of normal priority messages from servers to be processed in an sdmgr_im processing cycle.
Default: 1
Specifies the maximum number of notifications to process in an sdmgr_im processing cycle.
Default: 20
Controls the frequency at which you want to perform the cache cleanup for the Replication Manager.
Default: 60
Limits: 0 to 10080
Controls the frequency at which you want to perform the cache cleanup for the Task Manager.
Default: 30
Limits: 0 to 10080
Specifies the maximum time in seconds that an API server waits to gain access to a folder object that is currently occupied by another software delivery process. If the API server does not gain access within the specified time, it returns an error and the actual request fails.
Default: 0
With Maintenance: Asset report start hour, specifies a time period during which asset reporting is enabled (domain manager to enterprise manager).
Example: Asset reporting is enabled between 9:00 PM and 11:00 PM, that is, Maintenance: Asset report start hour = 21 and Maintenance: Asset report end hour = 23. At all other times asset reporting is disabled. This is useful on systems with large databases and many assets. On such configurations, asset reporting can take a long time and delivery of jobs is disabled during that period.
Default: empty
With the Maintenance: Asset report end hour policy, specifies a time period during which asset reporting is enabled (domain manager to enterprise manager).
Default: empty
Controls scanning for active computers in the LAN. If a value less than 0 is specified, the computer scan task runs every time the Task Manager runs. If set to 0, scanning is disabled. In all other cases, scanning takes place every 10th time the computer scan task runs. Hence, scanning takes place approximately every 150-300 seconds. If a value greater than 0 is specified, this value (in minutes) determines the interval between the times when the computer scan task runs.
Note: The Task: Scan sleep time policy controls the length of the Task Manager's sleep intervals. The length of the Task Manager activity intervals depends on the Software Delivery load.
This policy also controls two other functions:
In both cases, the job lists are by default checked every 480th time the computer scan task runs.
Default: -1, <locally managed>
Controls asset reporting. If False, assets are reported to the enterprise manager. If True, disables asset reporting, which can improve the performance of the Task Manager on a heavily used enterprise manager.
Default: False
Specifies in minutes how often the job cleanup task runs. Whether a job is removed depends on the value set for the Maintenance: Job cleanup time policy.
Limits: 0–65536
Default: 60, <locally managed>
Specifies the number of hours that a job that has executed successfully is displayed in a job container. Valid values are non-negative integers. The default is 240 hours (10 days), after which the job cleanup task runs. Renewed jobs are not covered by this cleanup; they remain.
Note: Setting the value to zero (0) could result in your not being able to see directly the result of a successful job, because it would be deleted immediately after execution.
Limits: 0–65536
Default: 240, <locally managed>
Specifies a limit for the deletion of obsolete jobs. If set to zero (0), all computer jobs that satisfy the obsolete condition (that is, the Maintenance: Job history cleanup time timer has elapsed) are removed. If the policy has a different positive integer value, a maximum of that number of computer jobs satisfying the obsolete condition are deleted.
Default: 0
Specifies which records in OK or error state are cleaned up after the Job history cleanup time timer has elapsed. Valid values are as follows:
Clean up all failed installation records for installed software and all records for software that is no longer installed/delivered.
Clean up all non-installation records and all failed installation records for installed software, and all records for software that is no longer installed/delivered.
Clean up all failed records for installed software and all records for software that is no longer installed/delivered.
Default: 2
Specifies the number of hours that a job that has executed is displayed in the Computer-Jobs folder. Valid values are non-negative integers. The default is 4320 hours (180 days), after which the job history cleanup task runs.
Note: After the time expires, neither installation jobs, if they refer to software that is still installed, nor delivery jobs, which refer to software still present in the scalability server library (SSLib), are deleted.
Limits: 0–65536
Default: 4320, <locally managed>
Controls how often an output file cleanup is performed. If set to 1, it runs every time the job history clean up runs (see Maintenance: Job history runtime). If set to 2, it runs every second job history cleanup; if set to 3, it runs every third job history cleanup, and so on.
Default: 1
Specifies the number of days that a job output file is maintained. The default is zero (0), which means the output file is maintained until the actual job is removed or, in the case of an installation or delivery record, never removed. Note that the file will not considered for removal until the actual job is about to be considered for cleanup (see Maintenance: Job history cleanup time).
Default: 0
Specifies a limit for the deletion of output files during a single cleanup. If set to zero (0), there is no limit.
Default: 0
Specifies the daily start time when the job history runs. The format is 24-hour HH:MM.
Examples: 02:00 for 2:00 a.m., 14:20 for 2:20 p.m.
Default: 02:00, <locally managed>
Controls the Installation Manager concurrency behavior. You can set the parameter value to True or False. When set to False, the Installation Manager runs as a single process and performs the notification processing and the database updates alternately. When set to True, the Installation Manager creates an additional manager process, Notification Manager, to handle notifications and messages related to the notifications.
Default: True
Controls the Task Manager concurrency behavior. You can set the parameter value to True or False. When set to False, the Task Manager runs as a single process and performs the tasks in a sequential manner. When set to True, the Task Manager creates additional manager processes, Policy Manager and Replication Manager, to perform the policy evaluation tasks and the replication tasks.
Default: True
Specifies whether to use the optimized transaction on the Installation Manager to update the job status messages and the software records in bulk. When set to True, the Installation Manager processes and completes the database update of the job status messages and the software records in bulk. When set to False, the Installation Manager processes and completes the database update of each message individually.
Default: True
Note: We recommend that you change the value of the Minimum Lock Percent policy only when you see a message about the unavailability of sufficient locks in the Installation Manager log.
Specifies the minimum percentage of job container and job locks to acquire before the Installation Manager can process the job status messages in bulk. The policy is applicable only for the bulk database updates of the job status messages. While processing the job status messages in bulk, if the Installation Manager gets less than 100% of locks but more than the minimum percentage of locks, the Installation Manager processes all the messages for which it has received the locks, and sends the remaining messages back to the Installation Manager queue. However, if the Installation Manager does not get the minimum percentage of locks, it sends all the received job status messages back to the Installation Manager queue for update in the next try.
Default: 50
Limits: 1 to 100
Determines the behavior of a domain manager when it receives a report from an agent that it has had a new OS installation. Valid values are as follows:
All installation records are marked uninstalled for the agent. No RAC automation is performed by the Software Delivery agent.
All installation records are marked uninstalled for the agent. An unsealed RAC job container is created for the Software Delivery agent containing all successfully executed jobs.
All installation records are marked uninstalled for the agent. A sealed and activated RAC job container is created for the agent containing all successfully executed jobs. Note that this option is recommended only if you have a clear picture of the list of successful installations for the computers and they do not involve unnecessary installation chains.
Default: 0, <locally managed>
Specifies whether the job linkage option, Ignore cascading install of dependent packages, should be set when a RAC container is created. If this policy is set to False, the jobs that are part of this job container when it is being evaluated are automatically cascaded. If True, the jobs are not automatically cascaded.
Default: False
Specifies whether the job option, Ignore job calendars on target computers, should be set when a RAC container is created. If this policy is set to False, calendars are not ignored at target computers. If True, calendars are ignored at target computers.
Default: True
Specifies the time period in hours that a RAC job order is valid after the date and time is set. When the timeout has expired, the job order will not be performed but remains visible in Software Delivery until it is removed. The range for this policy is by default 3–168 hours (as defined by the Software Job options: Minimum activation time and Software Job options: Maximum activation time policies). If, for example, the value is set to 1 hour, it is immediately replaced by 3 hours by the Software Delivery system.
Default: 168
Determines the behavior of a domain manager when it receives a report from an agent that it has had a new OS installation. If set to False, RAC is completely disabled and any old installation records are kept intact. If True, the RAC: Automation setting policy sets the behavior of RAC.
Default: True
Specifies whether a delivery job to a scalability server library is automatically set up, unless the delivery has already executed, when a RAC job is set up for an agent downstream from the scalability server. If False, a delivery job is not set up; if True, a delivery job is set up.
Default: False
Indicates whether it is permissible for other job containers to target the same computer during a computer job. To preserve the execution order of procedure dependencies, every non-fulfilled dependency must be cascaded as part of the same job container. Because of this, a computer job will fail during job container evaluation if there are any other job containers targeting the same computer with any of the procedures defined in the dependency list. After those computer jobs have completed, a subsequent reevaluation of the failed job container will succeed.
If True, a computer job will not fail in the scenario above, but may lead to situations where the execution order of procedure dependencies is not honored.
Default: False
Indicates whether the mixing of Software Managed (SM) products (SXP, PIF, PKG, and RPM) with other procedure types in batch job containers is enabled. If False, mixing is disallowed.
Note: This setting affects the Reinstall After Crash (RAC) behavior. If set to False, and RAC occurs for a system which does have a mix of Software Managed (SM) products and other procedure types and a mix of those is to be reinstalled, no RAC job container will be created. Depending on the RAC policy, the installation history for the system may get cleared.
Default: True
Indicates whether a prerequisite check for a container job is performed. If False, no prerequisite check is done.
Default: True
Indicates whether checks of software policies during job container evaluations are enabled. Valid values are True and False.
Example: If a software policy has a job assigned that ensures that the SD Catalog is not installed for the software policy group members, then a job that installs the SD Catalog will fail in the evaluation step for all targets that also belong to the template group.
Default: False, <locally managed>
Specifies whether to fail a job involving an SXP package whose cascaded procedures result in an implicit downgrade of the SXP package. If set to True, the corresponding job is set to the failed state and the implicit downgrade of the SXP package is not permitted. If set to False, the implicit downgrade is permitted.
Example: Consider an SXP package A 1.0/00 that depends on another SXP package B 1.0/00. The SXP package A 1.0/00 is then targeted to an agent that has already installed a higher version of the SXP package B, say SXP package B 2.0/00. If the configuration policy value is set to True, the implicit downgrade is not permitted and the corresponding job is set to the failed state. However, if the value is set to False, the implicit downgrade of the SXP package B is permitted on the agent.
Note: This configuration setting is specific to the packages of type SXP.
Enables an additional check for the need for DTS file transfers from the domain manager directly to agents connecting via scalability servers. If the specified software is already registered to the staging library of the corresponding scalability server, the DTS transfer will be allowed from there rather than from the software library of the domain manager.
Default: False, <locally managed>
Enables an additional check for the need of file transfers to scalability servers. If the Store Package in the Scalability Server Library option is selected, the file transfer initiates automatically. If this option is not selected and this policy is set to True, all jobs needing a file transfer to a scalability server will fail in the evaluation step of a container.
Default: False
Indicates whether the manager checks if a software item that has been detected as installed can be reinstalled. If True, a check is performed. If False, that is, no check, you will always be able to install the software item.
Default: False
Indicates whether an event is sent for every job completion on an individual agent. If True, job completion events are disabled. This can improve the performance of IM on a heavily used manager. If False, job completion events are enabled.
Default: False
Specifies the maximum number of hours you can designate for the give-up time of an item procedure in the Job Options dialog when setting up a job.
Default: 168, <locally managed>
Specifies the minimum number of hours you can designate for the give-up time of an item procedure in the Job Options dialog when setting up a job.
Default: 3, <locally managed>
Specifies how the main job status is to be displayed when the job succeeded on most computers but one or more computers had the product already installed. If set to OK, the main job status displays success, even though some of the computer jobs failed because they were already installed. Valid values are OK and NOK.
Default: OK, <locally managed>
Specifies how the registration status of an item from an enterprise manager is treated if the item is already registered locally. If set to OK, the main delivery status displays success even though some of the registrations failed because they were not unique. Valid options are: OK and NOK.
Default: OK, <locally managed>
Controls how software packages distributed from the enterprise manager are registered with the Software Library at the domain manager. If this policy is set to True and the package to be registered contains at least one Catalog-enabled procedure, the software package is automatically inserted into the Catalog folder for all users. If False, this behavior is switched off.
Note: If you change a procedure to be Catalog-enabled for a previously registered package that is without any Catalog-enabled procedures in the Software Library, that package will not be automatically inserted in the Catalog folder. You will have to do that manually.
Default: False
Specifies the computer evaluation refresh time interval in seconds.
Default: 60
Specifies the time in minutes between the time of the last agent creation or update and the actual evaluation of the on-demand software policy evaluation.
Default: 2
Specifies the Ad-Hoc task evaluation delay time in minutes for locked targets. When agents are locked by Move, Roam, or RAC, and are scheduled for on-demand software policy evaluation, the evaluation attempt is aborted, but a new attempt will be made later based upon the value of this policy setting.
Default: 30
Configures how many pending targets force an on-demand evaluation to start. Valid values are non-negative integers. If set to 1, each pending target computer is assigned its own evaluation task. Otherwise, each target registering within the 2-minute delay shares the evaluation task and postpones the activation of the task for another two minutes, until these two minutes pass or the limit specified here is reached.
Default: 10
Specifies when the Software Delivery system should invoke automated actions like the evaluation of Ad-Hoc Software Policies for preregistered computers, including those computers created by the Operating System Installation Management (OSIM) system. If False, any automated actions take place immediately after the computer has been preregistered, that is, before any real attributes and inventory have been collected by the engine. If True, the automated actions do not take place until the computer has been registered by the engine as part of a collect task.
Default: False
Switches the entire ad hoc, or on-demand, functionality on or off for all groups. If set to False, all on-demand functionality is disabled, overriding the on-demand functionality of the individual groups. If set to True, on-demand functionality is enabled.
Default: True
Defines the condition for skipping target systems during template group evaluation for a specific procedure if they were previously targeted by a job using the same procedure.
If set to True and another active job targeting the system is found during target group evaluation, the system will be skipped if any of the systems targeted by that job has not completed execution. If set to False and another active job targeting the system is found during target group evaluation, the system will be skipped only if the execution of the active job for that particular target system has not yet completed.
Default: True
Example: If set to True and you seal a template with two targets and one of the targets is offline (Target A) and the other target (Target B) fails during execution, a new container is not created for the failed jobs until either Target A comes online to process the procedures, the job times out, or the active job container is deleted.
Specifies the minimum time interval in seconds between evaluations of software policies.
Default: 0
Indicates whether a full evaluation should be performed when a Ad-Hoc group is sealed. If True, the full evaluation is performed. If False, full evaluation is disabled.
Default: True
Specifies the priority of the acknowledgement task running on a domain manager. Valid values are 1–10, with 1 as the highest priority.
Default: 5
Specifies the number of seconds the task manager sleeps between each pair of tasks. Historically, a value of 2 seconds was used on weak computers with a heavy load to prevent the Task Manager from stealing CPU and I/O time.
Default: 0
Specifies the priority of a catalog job task running on a domain manager. Valid values are 1–10, with 1 as the highest priority.
Default: 5
Specifies the priority of the computer evaluation (ad hoc) task running on a domain manager. Valid values are 1–10, with 1 as the highest priority.
Default: 5
Specifies the priority of the computer mover task running on a domain manager. Valid values are 1–10, with 1 as the highest priority.
Default: 5
Specifies the priority of the file transfer monitor task running on a domain manager or an enterprise manager. Valid values are 1–10, with 1 as the highest priority.
Default: 5
Specifies the minimum number of seconds a task of priority 1 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: 0
Specifies the minimum number of seconds a task of priority 2 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: 0
Specifies the minimum number of seconds a task of priority 3 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: 0
Specifies the minimum number of seconds a task of priority 4 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: 0
Specifies the minimum number of seconds a task of priority 5 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: -1
Specifies the minimum number of seconds a task of priority 6 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: -1
Specifies the minimum number of seconds a task of priority 7 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: -1
Specifies the minimum number of seconds a task of priority 8 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: -1
Specifies the minimum number of seconds a task of priority 9 has to wait until it is restarted before a lower priority task in the same time slot. A value of -1 means infinite delay, which causes the task to be started at most once in a time slot.
Default: -1
Reduces the number of notifications about task state transitions if set to 1 (True).
Default: 0 (False)
Specifies the priority of the maintenance task running on a domain manager or an enterprise manager. Valid values are 1–10.
Default: 5
Specifies the priority of a RAC job task running on a domain manager. Valid values are 1–10.
Default: 5
Specifies the priority of the receive task running on a domain manager. Valid values are 1–10.
Default: 3
Specifies the number of seconds the task manager sleeps between scans of the task list.
Default: 15
|
Copyright © 2013 CA.
All rights reserved.
|
|
