Jump to content

Informatica Experts... Question Regarding Purging


Recommended Posts

Posted

We run lots of continuous workflows so it is creating lots of entries in repository tables.

What is the best or recommended way to purge those entries? What i know right now is using repository manager.

Please let me know what informatica suggests or what people are usually doing.

Thanks

Posted

purging ante no idea. sorry2

Posted

naaku antha idea ledhu ee topic meedha... good post

Ganesh, yankees, google2011, okatonumberkurradu, tiger sathi, informatica86 inka evarunna vallu kooda.... answer please :15_3_321: :15_3_321:

Posted

[url="http://datawarehouse.ittoolbox.com/groups/technical-functional/informatica-l/deletion-of-objects-whose-versions-is-older-than-last-5-2875200"]http://datawarehouse.ittoolbox.com/groups/technical-functional/informatica-l/deletion-of-objects-whose-versions-is-older-than-last-5-2875200[/url]

Posted

Purges object versions from the repository database. You can purge versions of deleted objects and active objects. An object is a deleted object if the latest version is checked in and it has the version status Deleted. Other objects are active objects.
When you purge versions of deleted objects, you purge all versions. The deleted objects must be checked in. You can purge versions for all deleted objects or for objects deleted before a specified end time. You can specify the end time as a date and time, a date only, or a number of days before the current date.
When you purge versions of active objects, you can specify purge criteria. You can specify the number of versions to keep and purge the previous versions, and you can purge versions that are older than a specified purge cutoff time. You cannot purge a checked-out version or the latest checked-in version.
If you purge versions of a composite object, you need to consider which versions of the dependent objects are purged.
The PurgeVersion command uses the following syntax:
purgeversion
{-d <all | time_date | num_day> |
{-n <last_n_versions_to_keep> |
-t <time_date | num_day>}}
[-f <folder_name>]
[-q <query_name>]
[-o <output_file_name>]
[-p (preview purged objects only)]
[-b (verbose)]
[-c (check deployment group reference)]
[b]Note: [/b]You can use the -n option and the -t option together. They are not mutually exclusive.
The following table describes pmrep PurgeVersion options and arguments:

Posted

Ni log file ekkada unnayi ? windows looo naaa ? leka unix lo naaaa ? edaina third party program raasukoo... fox example windows lo unnayi anukoo ... edaina .Bat file raasukoo to delete the log files.. and call that .Bat file from informatica.. or from a third party tool.... Nenu ayithe alage implement chesa naa project looo :) good luck ...

Posted

we can write a shell script in unix to execute it every week or two week or monthly so that it will delete the files automatically. automate the job so that you dont have to do it manually everytime

Posted

what i understood is

when we run continuous workflows it will create lots of entries in repository tables

Foe example

If a workflow complete its run

OPB_TASK_INST_RUN table will be populated withn the following information
[SUBJECT_ID]
[WORKFLOW_ID]
[WORKFLOW_RUN_ID]
[WORKLET_RUN_ID]
[CHILD_RUN_ID]
[INSTANCE_ID]
[INSTANCE_NAME]
[TASK_ID]
[TASK_TYPE]
[START_TIME]
[END_TIME]
[RUN_ERR_CODE]
[RUN_ERR_MSG]
[RUN_STATUS_CODE]
[TASK_NAME]
[RUN_MODE]
[VERSION_NUMBER]
[SERVER_ID]
[SERVER_NAME]
[FRAGMENT_ID]
[SERVER_NODE_ID]
[SERVER_NODE_NAME]

and with the continuous run of the workflow, the table will grow enormously

right now what am doing is on every saterday i will run a script which will truncate that table

Posted

[quote name='jaffanajeffada' timestamp='1341935247' post='1302107803']
Purges object versions from the repository database. You can purge versions of deleted objects and active objects. An object is a deleted object if the latest version is checked in and it has the version status Deleted. Other objects are active objects.
When you purge versions of deleted objects, you purge all versions. The deleted objects must be checked in. You can purge versions for all deleted objects or for objects deleted before a specified end time. You can specify the end time as a date and time, a date only, or a number of days before the current date.
When you purge versions of active objects, you can specify purge criteria. You can specify the number of versions to keep and purge the previous versions, and you can purge versions that are older than a specified purge cutoff time. You cannot purge a checked-out version or the latest checked-in version.
If you purge versions of a composite object, you need to consider which versions of the dependent objects are purged.
The PurgeVersion command uses the following syntax:
purgeversion
{-d <all | time_date | num_day> |
{-n <last_n_versions_to_keep> |
-t <time_date | num_day>}}
[-f <folder_name>]
[-q <query_name>]
[-o <output_file_name>]
[-p (preview purged objects only)]
[-b (verbose)]
[-c (check deployment group reference)]
[b]Note: [/b]You can use the -n option and the -t option together. They are not mutually exclusive.
The following table describes pmrep PurgeVersion options and arguments:
[/quote]
[quote name='yankees' timestamp='1341939769' post='1302108173']
we can [b]write a shell script in unix to execute it every week or two week or monthly so that it will delete the files automatically[/b]. automate the job so that you dont have to do it manually everytime
[/quote]

But here i have a question
What happens at the time of executing the shell script and truncating the table AND the same time workflow is running and populating the same table

Posted

[quote name='informatica86' timestamp='1341940628' post='1302108242']
But here i have a question
What happens at the time of executing the shell script and truncating the table AND the same time workflow is running and populating the same table
[/quote]

Probably dead lock ayyi fail avutundhi anukunta... sCo_^Y sCo_^Y

×
×
  • Create New...