This document contains official content from the BMC Software Knowledge Base. It is automatically updated when the knowledge article is modified.
Control-M Workload Archiving
Control-M Workload Archiving
Control-M Workload Archiving
General questions for Control-M Workload Archiving
1. User "arcuser" , what permissions are needed in RACF in z/OS? The documentation mentions BIMUSER but it doesn't exist on z/OS.
RACF permissions should be similar to GCSERV user and should have at least permission to view log and view sysout for Workload Archiving to get the output and the logs of the jobs.
2. How can we perform a space calculation to figure out what we need in disk space?
There is no real formula or number. It depends on the customer data (average output size) and the number of jobs you would want to archive.
This is something that is specific for each customer since it depends on the rules, size of log and output, retention policy and database type.
There is some compression on the database (30% to 50%) and it also depends on the nature of the output of jobs.
What we suggest regarding the calculations is to monitor the amount of data after a week and make the calculations based on those numbers taking into account the retention period.
This can be more controllable when setting the maximum output size (in CCM->Tools->Workload Archiving->Archive Policy) and the retention period (default is 7 days and can be set in each rule).
3. Does it has a maximum process to treat? Here in production we have over 1 million runs.
We recommend using rules to archive the important data, not all of it.
We've made tests using small files (1-10 Kb) on LAN connection. The application supported collection rate of 12 output per second (around 1 million a day) using the
default configuration. In any case there are ways to configure the system to increase the collection rate (so more jobs will be collected, faster).
4. Is it possible to have the database schema?
The database schema is not part of the Control-M/EM database schema since it is a separate database so separate schema will have to be created.
5. What is the meaning of the following repetitive messages from the archive.log:
-rw-r----- 1 ctmarchv ctmarchv 9022363 2015-01-14 10:28 archive.30950.20150109.0.log
0114_10:27:45.740 [INFO] T@24-Clct-free too many items are waiting to be collected, not processing new items. # waiting items:1127 CollectionManager::run
0114_10:27:47.741 [INFO] T@24-Clct-free too many items are waiting to be collected, not processing new items. # waiting items:1127 CollectionManager::run
0114_10:27:49.741 [INFO] T@24-Clct-free too many items are waiting to be collected, not processing new items. # waiting items:1127 CollectionManager::run
0114_10:27:51.742 [INFO] T@24-Clct-free too many items are waiting to be collected, not processing new items. # waiting items:1127 CollectionManager::run
This is an informative message indicating there are many jobs waiting to be collected.
The data is not lost. The jobs are "waiting" in the DB (first 1000 are waiting in memory, the rest are in the DB)
6. If 2 Archive rules exist for a single job, which one will apply? Example: Rule 1 keeps Job logs and output for 1 month, and Rule 2 keeps Job Logs and output for 1 year.
The first rule that applies to the job will be the rule used. Any additional rules will not apply to the job. Please also note 13 (below).
7. What compression algorithms are used to store job output.
For Oracle there is oracle advanced compression which is available with Enterprise edition of Oracle. It should be enabled by the DB admin and we would use it.
It will compress the logs and output. We allow working even without it in order to avoid making it as a prerequisite as there could be sites that don't have it available. Internal tests show a compression ration of 3:1.
Same with PG. the compression used is provided by the PG package and we see a compression ration of 2:1. The compression results are data specific and each environment will yield different results.
8. What is the limit of each job output size.
Default is set to 10 Mb,
It can be configured up to 100 Mb.
9. What are the expected results of deactivating a policy.
When a policy is deactivated all the data remains in the repository and will be purged based on policy retention period no matter if rule is active or not active rules.
10. What are the expected results when a rule is deleted.
When you delete a rule (not disable it) there is a pop up asking if you want to delete the related data. Related data will be deleted only if user response Yes.
11. Is v19 release or WA support PosgresSQL upgrade procedure
No, with v20 the customer will be able to use external PG
12. When doing a install for Control-M Workload Archiving , you can choose a small or large install and either an Oracle or PostgreSQL database.
What are the database requirements for each?
For Postgresql the difference is within the shared_buffers and max_connections parameters:
shared_buffers = 2000 MB
max_connections = 1200
shared_buffers = 500 MB
max_connections = 200
In Oracle the difference is in the tablespace configuration and size
Size = 5000 MB
Size = 1000 MB
No data compression
For production environment please use only the Large configuration.
13. How are the rules evaluated for collecting archived data
Before v20 only the first rule that the job satisfied fires. The rules are evaluated in the order of definition.
With release of v20 and higher every eligible rule that meets the criteria of the job will fire.
14. What ports need to open for Archiving to work if case a firewall if enabled
- Ports from distributed archive GSR to primary GTWs
- Port from distributed archive to primary CMS
- Port from distributed archive to primary GSR (not default configuration but a possible scenario)
- Port from primary EM (GSR, CMS or Webserver) to distributed archive
- Before v18 - Port from the client directly to the distributed archive
- Port from distributed archive to external archive DB - external DB host and port are defined by the customer and should not be blocked
The Control-M Workload Archiving Server portion is not in HA mode;however when failover to the Secondary, Control-M Workload Archiving Server will continue to work and the Secondary will connect to the
Control-M Workload Archiving Server instead of the Primary.This is because Control-M Workload Archiving Server should be installed only on "Additional distributed installation".
If the Control-M Workload Archiving Server component in Control-M Configuration Manager (CCM) is up and running, It must be connected to the active EM (secondary in a case of failover) because all EM components on the inactive EM
(primary in a case of failover) are down.This can verify by running ctm_em\archive\script>arc_test_configuration (on the distributed machine).