Data Server Manager stores very useful information in its historical repository database that you can directly access through SQL. You can run the following statement against a DSM repository database to see the details of key metrics across all the databases monitored by DSM. (There is more information on accessing this information in the DSM Wiki.)
WEEK(COLLECTED) AS WEEK_COLLECTED,
SUBSTR(DBCONN_ID,1,14) AS DBCONN_ID,
SUBSTR(WORKLOAD_NAME,1,14) AS WORKLOAD_NAME,
DATE(MAX(COLLECTED)) AS END_DATE,
SUM(TOTAL_CPU_USEC_DELTA) / 1000000.0 AS CPU_SEC,
SUM(LOGICAL_READS_DELTA) AS LOGICAL_READS,
SUM(ACT_COMPLETED_TOTAL_DELTA) AS ACTIVITIES,
SUM(TOTAL_APP_COMMITS_DELTA) AS COMMITS
GROUP BY WEEK(COLLECTED), DBCONN_ID, WORKLOAD_NAME;
Try running it in the DSM SQL Editor
You can run this SQL directly in the DSM SQL Editor and export the results to an Excel file and easily pivot on the data.
Work like a Data Scientist
If you want to mine the data like a Data Scientist, try accessing the same data through a Data Scientist (Python) notebook.
The minedsmhistory notebook file I stored in Github does all the work for you. It retrieves data from the Data Server Manager monitoring history and maps out resource usages across your Db2 Enterprise. (You will need connection privileges to the Data Server Manager repository database.) It creates a view in the repository database that provides a list of databases and workloads in each database for each data point collected. The values collected are key resource metrics: CPU seconds, Transactions, and Logical Reads. It then maps out a pivot table for the total resources or throughput per database and workload by week for each database in your Data Server Manager monitoring enterprise.
Using with DSX Desktop
By using IBM DSX (Data Scientist Experience) desktop you can work like a Data Scientist in minutes. Start by downloading the DSX Desktop from IBM. Once you have installed the DSX desktop you can create a new notebook by selecting My Notebooks and New Notebook and From File. Download the MineRepository.ipynb file from this project and choose the file from in the Notebook File section of the Create Notebook form. Provide a new notebook and select Create Notebook. You can then run the notebook by selecting the run cell icon for each step in the notebook. (Make sure you update cell number three with your unique database connection information.)
Try it out and let me know what you think, especially if you create your own notebook that you would like to share.