-killsession <Session ID>
You use this parameter to log off the session of the current user. The parameter logs off only the user interface sessions. You cannot log off sessions that were started using command line programs (runppmconfig, runppmimport, for example).
-showusers
Outputs all sessions for the users currently logged into the system to the console in tabular form.
-clearcache
Clears the caches used in the system.
-fillcache
Triggers the calculation of the caches used in the system. Only entries that are pending recalculation will be updated. If you want to make sure all favorites are recalculated, call runppmadmin -clearcache before.
If you have deactivated automatic filling of the cache in the client-specific file Keyindicator_settings.properties with the setting AUTO_FILL_CACHE=false, each time data relevant to the cache changes, you can trigger re-filling of the cache using the command line with the –clearcache and -fillcache options.
-clearpchistory
The entire Performance Dashboard history cache will be deleted.
-savepchistory
The contents of the memory-based Performance Dashboard history cache are saved in the database. Already existing cache content will be lost.
-addpchistory
The contents of the memory-based Performance Dashboard history cache is extended by the newly added entries. Already existing cache entries will be retained.
-memoryinfo <file>
The following information is output and written to the file specified:
This information is intended for experiences users to help identify optimization potential of the analysis server.
When executing this command, you may experience wait times in the analysis because the analysis server first runs an internal memory streamlining process (garbage collection) before collecting the required data. During this time, the server does not respond to analysis queries.
Selectivity of configured indices
For each index the average selectivity regarding the queries is displayed.
The selectivity is calculated individually for each request in which the dimension appears as a filter. The selectivity is calculated as follows:
Expected number of process cube rows (process instances) to be visited if the specified filter values of this dimension are used, divided by the number of total rows if only the process type filter is used.
In the index statistics, the average of the selectivity of the individual requests per dimension is specified. In addition, the selectivity values of the dimension < 25%, <50%, <75%, and <100% are displayed. A small selectivity close to 0 (or 0%) is good, because this means that by using the index only very few data rows/process instances had to be checked. Without index, considerably more data rows would have had to be checked.
-recoveryfile force
With the -recoveryfile force option, you force the recreation of recovery files of the PPM analysis server, regardless of server status. For each client, all recovery files are created.
By default, the recovery files of the client are saved under <installation directory>\ppm\server\bin\work\data_ppm\recovery\<client>.
The path to the recovery files is defined by the key RECOVERY_FOLDER in the client-specific file AnalysisServer_settings.properties.
-statisticinfo <file>
Returns statistical information as output and writes the output to the specified file. In scaled systems, the statistical information of each analysis (sub-) server is displayed separately.
The following information is returned as output:
The output for the first four points is identical to the output of the -memoryinfo parameter. This parameter differs from the -memoryinfo parameter because it does not instruct the server to run garbage collection and the output does not include the information about the runtime environment. The information on the configured indices also includes statistical information on the indices.
The statistical information on the queries is also included in the output for the analysis engine for all process dimensions for which an index can be created:
- Number of all queries to the analysis server.
- Number of all queries to the analysis server with a filter on a process dimension,for which the server can create an index.
- Table of queries on process dimensions (the first four columns together form the primary key of the table):
To calculate the selectivity, PPM divides the number of process cube lines that correspond to the specified process type filter by the number of viewed/visited process cube lines. A high selectivity, that is close to the value of 1 (or 100%) is good, because it indicates that only very few of the visited rows were not part of the result.
Note that only queries that are sent to the analysis engine are included in the statistical information. Queries that use data from the cache are not included in the statistics. The statistical information in not stored and all statistics are deleted after a restart of the analysis server.
-clearstatistic
The statistics information of the analysis server is deleted. In scaled systems, the statistics information of all analysis (sub-) servers is deleted.