Sudden growth of DPA repository, which instance caused the growth?


We monitor 4 instances, our DPA repository had grown and we added diskspace and truncated logs, is there a way a query against the tables on the DPA repository to see which instance caused the growth?

Dennis (SAIC)

  • Hi

    Regardless from the database system used as the repository you should be able to identify which tables are taking up the most space.

    As soon as you have collected the information you can go a step further.

    Most of the repository table names does looks like <table_name>_<id>.

    E.g. CONACT_1

    You can then query the repository table COND which contains the monitored instance name plus the id. The id column is matching the id in the table name.

    SELECT id, name from COND;



  • Hi Fabian,

    Thanks, I'm able to see all the tables by issuing: 



    I can also see the instances being monitored by issuing:

    SELECT id, name from COND;

    Which of my monitored instances relate to the growth of the transaction logs of the repository?  which of these tables can I query?  


Reply Children
  • Hi Dennis

    You could check now the size of the tables. There are a few options to get this. One of the option is via the DPA GUI:

    "Options" / "Support" / "DPA diagnostics" / "Tables". 

    Then you can sort by size or rows to get the biggest table.

    Link the ID in the table name with the ID column from the table COND. At least you know then which of the four monitored instances is related with the biggest table(s).

    Or are you looking for something else?



  • Hi Fabian, 

    I’m able to see [CONV_METRIC_DETAIL_545] and it's 23 gb in size and 540 million rows

     Queried [CONV_METRIC_DETAIL_545] and it has the below:

     I've attached a word do but were did it go? this has the screenshots.  Sorry, I'm new to THWACK.

    I issued:

      SELECT id, name from COND;

     And it has ID and name of instances being monitored

    How do you link this  [CONV_METRIC_DETAIL_545]?  Where is the common column so I can join it?