SAP Library - Find your answers

Use

Collection and reorganization schemata specify how long data assigned to this schema are to be stored in the Central Performance History (CPH) with which resolution. Predefined schemata are delivered with the CPH. These fulfill most requirements. You have two options for creating new schemata, or displaying or changing existing schemata:

· Use the expert setting (see Maintaining Collection and Reorganization Schemata (Expert Setting)).

· Use the standard setting described here.

In the standard setting, you cannot create new schemata from scratch. Instead, you can use an existing schema as a template for your own schema, by changing the storage period of the various aggregates. If you want to create your own schema, we recommend that you use one of the predefined SAP schemata as a template. You can recognize these templates, as they always have the prefix SAP. You will only have to use the expert setting in rare cases.

Procedure

...

1. You can call the standard setting for maintaining collection and reorganization schemata from several functions:

¡ In the overview screen of the CPH (transaction RZ23N), choose, in the Data Collection and Reorganization group box, the Maintain Schema pushbutton.

¡ On the Assign Collection and Reorganization Schemata screen (see Assigning Collection and Reorganization Schemata to Performance Values), choose the desired schema in the Collection/Reorganization Schemata subscreen, and then choose Display Schema ().

¡ In the Alert Monitor, with the maintenance functions activated, select a node, and choose Collect MTE Class in Central Performance History(). A screen displaying a list of the existing schemata appears. To display one of these schemata, select the schema and choose Display Details ().

2. In all of these cases, the Display Collection and Reorganization Schema screen appears. On this screen, you can see which schemata exist. If you want to edit a schema or to create a new schema on the basis of an existing schema, choose the desired schema by double clicking it, and choose Display « Change ().

3. The screen now has the following structure:

Creating or Editing a Collection and Reorganization Schema

4. At the left side of the screen, there is a list of existing schemata. The system displays the name of the selected schema in the Selected Schema field. If you want to use this schema as a template for your own new schema, enter the desired name in this input field.

5. In the Collection and Reorganization Process group box, the system displays the resolution of the data that is collected, and when it is reorganized into which resolution. If you want to change one of these times, choose Change Time (). The system displays a separate dialog box, in which you can enter the new period.

6. In the Advanced Settings group box, you can specify the time zone for the collection or the reorganization of the data and a day and calendar schema. For more information, see Aggregation, Day Schema, and Calendar Schema).

7. Save your entries.

Delete a Collection and Reorganization Schema

8. Select the desired schema and choose Delete Schema ().

9. Confirm the command.

If there are assignments for this schema, the system informs you of this. If you delete the schema despite this warning, the MTE classes that are assigned to this schema are no longer collected and are deleted from the CPH during the next run of the reorganization job.

See also:

Delivered Collection and Reorganization Schemata

Central Performance History of the Monitoring Architecture start page

Use

Collection and reorganization schemata specify how long data assigned to this schema (see Assigning Collection and Reorganization Schemata to Performance Values) is to be stored in the Central Performance History (CPH) with which resolution.

Use these expert settings if the normal screen for maintaining collection and reorganization schemata (see Maintaining Collection and Reorganization Schemata (Standard Setting)) does not meet your needs.

In the CPH, you can define in great detail which performance values you want to store from which systems in which resolution, and for how long. This number of configuration settings is, however, only necessary for a very small number of users, as SAP already delivers powerful standard schemata.

Prerequisites

The expert setting, in which you can set all parameters of the schemata as you wish, is described here. The prerequisites for using the expert setting are:

· Knowledge about the mechanisms of aggregation of data; you should understand the meaning of terms such as “Monthly aggregate with an hourly resolution”.

· Knowledge about the memory table of a Collection and Reorganization Schema, in which every row of the table determines the storage period of an aggregate with a certain resolution

Procedure

...

1. To call the expert setting for editing collection and reorganization schemata, on the normal screen for maintaining collection and reorganization schemata, choose Edit ® Screen for extended configuration.

2. The Display Collection and Reorganization Schema screen appears. On this screen, you can make all settings for existing schemata and create any new schemata that you wish.

Creating a Collection and Reorganization Schema

3. Choose the Create New Schema pushbutton () and enter a name for the collection and reorganization schema in the Name field.

4. Depending on the temporal resolution in which performance values are to be collected in the CPH (usually Minutes, Quarter Hours, or Hours), enter in the Day Aggregate line the time in days for which you want to retain this data.

5. After this time has elapsed, the corresponding data is deleted. However, performance values should still exist after this (although as a different aggregate and/or with a different resolution. Enter the storage periods for the desired combinations of aggregate and resolution in the table.

Note that the time unit of the storage periods is the unit of the corresponding aggregate in each case. This means that different units apply for the different rows of the table.

6. A collection and reorganization schema also includes the selection of the method with which performance data from remote systems reaches the CPH. In the Collection Method group box, choose Collection Method. For an explanation of the various methods, see Collecting and Reorganizing Performance Values.

7. Performance values are first stored in the local time zone in the monitoring segment. If necessary, you can convert these values to a different time zone either during the data collection or during the aggregation. This means that the workload that is created by changing the time zone is shifted to a less performance-critical part of the process. If you do not specify a time zone, the data is converted to the local time zone of the CPH system.

8. You can use a day schema and a calendar schema to specify whether particular hours or days should not be considered during the aggregation (see Creating and Editing a Day Schema and Creating and Editing a Calendar Schema). If you do not specify a day and calendar schema, all hours and days are used in the calculation of average values.

9. Save your entries.

Editing an Existing Collection and Reorganization Schema

10. In the Name group box, choose the selection help, and choose the collection and reorganization schema that you want to edit. Alternatively, you can also enter the desired schema in the input field and confirm the entry with the Enter key.

11. Switch to change mode by choosing the Display « Change () pushbutton.

12. Edit the schema; you can find a description of the individual settings in the previous section.

13. Save your entries.

Deleting a Collection and Reorganization Schema

14. In the Name group box, choose the selection help, and choose the collection and reorganization schema that you want to delete.

15. Choose Delete Schema () and confirm your selection.

Central Performance History of the Monitoring Architecture start page

Archiving and Reorganization are totally different issues

Archiving is a process when you archive old master/transactional records which are not in use any more from long time. You archive those records keep on some backup device. These records can be displayed, and only Display, can never be reloaded in tables again physically.
This will free up the space in tables from those records which are not in use any more but exist in tables physically.

Reorganization basically removes deleted data from tables. During different transactions you delete some records, which are deleted virtually but not physically. Even Archive process delete those records virtually.
So you run Reorganization to free up the space consumed by deleted records.

This make the difference between the two clear.

Now about reorganization.

It depends on which OS and DB you are using.

If Oracle, SAPDBA is best tool to reorganise.

In addition to above, you can reorganise the whole database, but it takes time depending upon your database. Some times it can be more than a day or you can reorganise a table suggested in Early Watch system.

It all depends on you.

Archiving is a process when you archive old master/transactional records which are not in use any more from long time. You archive those records keep on some backup device. These records can be displayed, and only Display, can never be reloaded in tables again physically.
This will free up the space in tables from those records which are not in use any more but exist in tables physically.

Reorganization basically removes deleted data from tables. During different transactions you delete some records, which are deleted virtually but not physically. Even Archive process delete those records virtually.
So you run Reorganization to free up the space consumed by deleted records.

This make the difference between the two clear.

Now about reorganization.

It depends on which OS and DB you are using.

If Oracle, SAPDBA is best tool to reorganise.

In addition to above, you can reorganise the whole database, but it takes time depending upon your database. Some times it can be more than a day or you can reorganise a table suggested in Early Watch system.

You have ARCHIVED the BSIS table (because of old data etc) and now this BSIS table has "holes" in physical structure and it's quite a large table.

Here are the steps to reorg. a single object:

1. Take a backup of your Oracle database.
2. Shutdown SAP
3. Start Oracle Database with Rollback Segment PSAPROLLBIG (all other smaller segments not started). Remember, the size of PSAPROLLBIG should be equal or larger than your BSIS table.
4. Ensure sapreorg directory has enough space for taking BSIS Oracle Export dump. The Oracle Export utility compresses the file to some extent.
5. Fire your SAPDBA from Unix level.
6. Go into SAPDBA "reorganisation" menu.
7. Select "Reorganize Single Table / Index"
8. Give BSIS as table name.
9. Start.
10. Don't Forget, first upgrade your SAPDBA tools first from OSS/SAPNet.
These include: 1. sapdba 2. brtools 3. brarchive 4. brbackup 5. brrestore 6. brconnect.

Evaluating Online Re-Organization Tools

Content Author: Christophe Rabeau

We are evaluating online reorg tools for an Oracle SAP database which is soon to be of 1 Terra Bite size. I would like some guidance for online reorg.tools and experiences of using some of them.

I use SAP 4.6C under Oracle 9.2 under AIX 5.1 on Risc6000 medium-performance machines.

I experienced some reorganizations SAP online, due to the PRD system that must always be up.

Fisrt, consider that the entire reorg is not possible when SAP is online, because of permanent tables (system tables and so on) that could not be exported/imported while running.

An entire TABLESPACE reorganization is no more suitable because of the long runtime between 2 offline backups.

So you can reorg only some tables/indexes or single objects. Be sure that the objects should not be accessed during the reorg, so it will fail. For a particular case, for example a BIG table, I experienced serious perf problems :

JVTLFZUO table : 10 Gb data + 17 GB indexes = 37 GB to be reorganized.

When SAP is online, it took about 13 hours (4 procs, 8 GB RAM)
When SAP is not available for users, it took less than 5 hours !!!

So my advice is :
- don't reorg too big tables with SAP online.
- don't reorg system tables with SAP online.
- you'd better to perform a reorg by night or in a lower user activity.
- you'd rather to reorg with SAP down for perfs point of views.
- always perform an OFFLINE backup before reorg, because the reorg is not always successful.