Considerations for splitting MDB Service Desk tables in SQL Server DBMS over multiple disks.
search cancel

Considerations for splitting MDB Service Desk tables in SQL Server DBMS over multiple disks.


Article ID: 20816


Updated On:


CA IT Asset Manager CA Software Asset Manager (CA SAM) ASSET PORTFOLIO MGMT- SERVER SUPPORT AUTOMATION- SERVER CA Service Desk Manager - Unified Self Service CA Service Desk Manager CA Service Management - Asset Portfolio Management CA Service Management - Service Desk Manager



This article provides some considerations for distributing MDB tables across multiple disks. This article applies to SQL Server.


The following actions may help with the performance of the MDB on SQL Server for the CA Service Desk Manager application:

  • Store MDB data and log directories on separate disk drives.

  • Configure TEMPDB to reside on disk separate from the disks containing the MDB.

  • Configure TEMPDB size to be large enough to contains the largest table expected to be stored in the MDB; set the incremental growth percentage to a size that prevents too much incremental growth but also a size that is not unnecessarily large.

  • Configure the SQL data to reside on striped disks as a preference to separate log/data disks.

  • Set up the transaction log files to cross multiple devices with a set incremental amount. The MDB transaction log incremental is 10% and unrestricted - you could set this to a specific amount rather than to a percentage.

Keep in mind that CA Service Desk Manager operations and functionality are based on objects. It may be necessary to understand the mapping of tables to objects. In an out-of-the box Service Desk, the act_log and evt_log tables are written to frequently.

To achieve the highest level of performance, the frequency, urgency, and importance of each operation within CA Service Desk Manger in relation to the needs of your business may need to be well understood. Measurements, projections, and adaptations may be necessary. Despite conclusions made after experimentation, there may be situation in which the results are not as expected..

Additionally, you may want to review the article titled "Configuring Database Files for Optimal Performance" which currently exists, at the time of this writing, at the following link:


Component: ARGIS