Frequently asked questions or issues experienced by CA Datafinder users

Document created by Sameer_Chandragiri Employee on Jul 24, 2014Last modified by Melanie_Giuliani on Nov 8, 2017
Version 4Show Document
  • View in full screen mode

Below are the most common questions or issues that Datafinder (aka Grid-Tools) users experience.


1.     My Datamaker/GT Subset/GT SDM License has expired and I would like to obtain a new license key. How can I apply for and obtain one?

a.      Refer to the document “LISA KB203 - Licensing CA LISA Datafinder that outlines the steps to apply for and obtain license keys for the various Datafinder products.


2.     When I run a masking job through GT SDM, the maximum number of records that are processed for any table is 5000. Why does this happen and how can I fix this?

a.      This indicates that there is a problem with the GT SDM license. By default, without a valid license, SDM will process only 5000 records at most per table.

b.      Make sure that the license you have been provided with has not expired and has been applied correctly. Refer to the section “Licensing process for SDM (Simple Data Masking)” in the document “LISA KB203 - Licensing CA LISA Datafinder that outlines the steps to apply the license key on your installation of GT SDM.


3.     How can I update my version of Datafinder (Datamaker)?

a.     Download the latest Datamaker files from the path : -

b.      Close all existing instances of Datamaker on your Datafinder Application server.

c.      Unzip the patch into your Datamaker installation folder, for example, "C:\Program Files (x86)\Grid-Tools\GTDatamaker" and overwrite the common files.

d.     Right click on the “GT Datamaker” icon and click "Run as administrator" and login to DM with username = administrator and the corresponding password.

e.     Press ALT + CTRL + M to bring up login screen of “Repository maintenance”. Enter the same Datamaker credentials as above.

f.       Expand the first tree item on the top (GT Datamaker Test Data Repository) and click “Check Repository Tables”. Click “Process” on the right pane. If the script finds problems, click “Execute Script” button to fix them.

g.      To verify that the scripts were updated, click on the “Update Repository PL/SQL” entry under the first tree item and it should say “XX attempted and 0 failed”.


4.     Where can I find the tnsnames.ora file on my Windows machine? (I am using an Oracle repository for use with Datafinder)

a.                            a. The tnsnames.ora file is usually found in the “ORACLE_HOME\network\admin” directory. For example: “C:\oraclexe\app\oracle\product\11.2.0\server\network\ADMIN”.


5.     SDM Masking performance is extremely slow.

a.      There could be a number of factors that could cause masking performance to be unsatisfactory. Some of them are:

i.     SDM Application Server and the target DB/Seed list DB/Cross-reference DB are not on the same LAN segment and the network latency could be causing performance delays.

ii.     Masking rules are not efficient – There may be more efficient ways of defining some of the masking rules. (For example, some Cross-reference rules may be replaced by HASHLOV).

iii.     There are Update triggers defined on the table(s) being masked.

b.      It is important to involve the customer site DBA to help determine whether the root cause is DB-related, such as triggers, full-table scans, etc.

c.      Once all the network/DB-related factors and masking rule inefficiencies have been addressed and resolved, if there is still a performance issue, then open a Support ticket and the issue will be addressed accordingly.


6.     I have access to the live test environment data only for 3 days, and I would like to make sure that I can successfully implement a masking solution on the data within that timeframe.

a.      First-time implementation of masking is almost always an iterative process. Live test environments CANNOT be blocked for masking implementations. The best practice is to set up a separate sandbox with a copy of the data for masking, which no other application teams are dependent on for immediate testing. Masking implementation needs to be first stabilized using a few clean runs and then eventually applied on a live test environment over a weekend.


7.     I have an issue with a specific SDM masking function. It is not working as expected.

a.      First, make sure you have referred to the GTSDM documentation for that specific function and are using the correct syntax as per the documentation. (You can find the documentation at “<GTSDM installation folder>/doc/GTSDM.pdf”).

b.      Send CA Support ALL the following files in order for us to reproduce and investigate the problem further:

i.     GTSDM Log file

ii.     Connection file (connectXXX.txt),

iii.     Mapping CSV file (that contains the masking functions),

iv.     Options file,

v.     DDL of the table that the issue is seen for,

vi.     Seed table connection file (if applicable to the scenario), and

vii.     Cross-reference connection file (if applicable to the scenario).


8.     I am using GT Mapper for SDM masking. I am using 20 masking rules and about 10 seed tables. I am getting an Out-of-memory error and also seeing other unexpected symptoms.

a.      GT Mapper is primarily meant to only support very small masking jobs mainly for demo purposes (say, 5-10 rules with no more than 2-3 seed tables). For any production-level jobs, always use the command-line GTSDM with increased memory allocated to the process.

b.      The best practice is to create the masking maps (CSV) through Datamaker (Transformation Maps window) and then modifying them either through the same screen or a text editor such as Notepad++. Then the masking job can be run from the command-line using the GTSDM command.

c.      Example syntax to run the job through command-line:

java -Xms100M -Xmx1000M -jar gtsdm.jar connectTravel.txt masking_map.CSV OPTIONS.txt

This will allocate 1GB of maximum memory for the GTSDM job.

d.      For more information, refer to the section “Managing Large Table updates or large seed tables” of the GTSDM documentation (You can find the documentation at “<GTSDM installation folder>/doc/GTSDM.pdf”).


9.     The GTSDM program hangs in the middle of the masking process. What could be the likely cause and how can I fix it?

a.      A GTSDM hang would most likely occur for one of the following reasons:

i.     A Null pointer exception in GTSDM: This should be captured in the command prompt or log file (An error log may not be produced in this case).

ii.     Out of memory error: This can be fixed by allocating more memory to the GTSDM process through the “–Xmx” option of the java command. For example:

java -Xms100M -Xmx1000M -jar gtsdm.jar connectTravel.txt masking_map.CSV OPTIONS.txt

This will allocate 1GB of maximum memory for the GTSDM job.

iii.     Table locks: The table to be masked is potentially locked by another process. No error messages would be produced in this case and GTSDM would just hang. To avoid table locks, make sure to commit all the DML transactions against the tables to be masked. You may also have to contact your DBA for assistance to ensure that none of the candidate tables are locked.


10.  GT Subset is not showing relationships that I defined in Datamaker.

a.      By default, GT Subset shows only FK relationships and any user-defined ones that were defined in GT Subset or copied from Datamaker.

b.      To copy relationships between Datamaker and GT Subset use the "Utilities -->
Copy Relationships" from GT Subset main menu. For more details, refer to
section "20.2 UTILITIES -> COPY RELATIONSHIPS TO AND FROM GT DATAMAKER" of your GT Subset user guide which you can find in your “<Datafinder installation folder>/doc” directory.


11.  While using GT Subset, I tried using Alias names in my driving table SQL but I am either getting errors or seeing unexpected results.

     a.    Alias names are not supported in GT Subset. This is because the internal SQL generator creates its own internal alias names in order to reference the correct table further along joins (when it builds
             its  join SQL for the extracts). This will conflict with the user-defined alias names.
1 person found this helpful