DBShadow for MaxDB Databases
|Libelle DBShadow® for MaxDB is a turnkey solution for setting up and managing MaxDB standby databases. It provides out-of-the-box support for MaxDB on all common UNIX, Linux, and Windows platforms. |
We're here to help!
|DBShadow® is part of the BusinessShadow® suite of products to mirror SAP landscapes which includes additional features to mirror flat files (FSShadow®), built-in IP failover components (SwitchApplication) and a common central Graphical User Interface.||Request a Call-Back|
MaxDB as Strategic Platform for SAP Applications
MaxDB is fully controlled by SAP AG and therefore probably the most strategic among the five databases supported for SAP database platforms. Some application components are solely supported on MaxDB, in particular LiveCache as part of the Supply Chain Management solution. A growing number of customers run their complete NetWeaver stack on MaxDB supporting any SAP component such ECC, BI, XI, Portals, and more.
Libelle DBShadow for MaxDB
The first version of DBShadow for MaxDB made its debut in 2003 and was continuously improved ever since. Libelle Development has been enhancing its functionality from two different angles: (1) Improvements specifically for MaxDB customers and geared towards MaxDB features such as interaction with the MaxDB database instance, handling of log volumes, medium management, and so on. (2) Improvements along with the BusinessShadow suite of products with functionality for all database platforms such as server agent management, wide area network tuning, file handling, and more.
Overall, the current version of DBShadow for MaxDB is years ahead of what customers are getting in functionality, stability, and manageability in comparison to developing custom scripts with reasonable efforts based on native MaxDB standby capabilities.
Setting up MaxDB Standby Databases with Libelle
DBShadow for MaxDB runs directly on the Primary System (MaxDB Production) and Secondary System (MaxDB Standby). Core server agents handle all tasks necessary to setup and manage the Standby System. Once installed and configured, these agents establish a communication link between Primary and Secondary system and interact with the MaxDB database instance on each server. Agents are then also able to communicate with an independent GUI for setting up and managing the configuration. Please find a screenshot of a basic configuration below.
To establish a standby database, DBShadow provides four distinct procedures which are configured and managed by Libelle software:
(1) ‘Initial Copy’ Process
An automated procedure to create a copy of the MaxDB database from the Primary to the Secondary System.
(2) ‘Archiver’ Process
Procedure to backup the MaxDB Log Volume in certain intervals (e.g. every 10 minutes) and transferring these backups / log files to the Secondary System.
(3) ‘Structure’ Process
Procedure to monitor database structure changes and update the same on the Secondary System
(4) ‘Recover’ Process
Procedure to manage the standby database during normal operation (continuously apply log files to the standby database) and to provide four distinct automated failover options in case the Standby database needs to be activated. The screenshot below shows the core options for Initial Copy, Archiver, and Recover Process. If you are interested in more details, Libelle Consultants are available for web-based live demonstrations any time.
Managing the Standby Databases using Libelle GUI
Once configured and with the ‘Initial Copy’ process executed, the DBShadow server agents on the Primary System are continuously running the ‘Archiver’ process for backing up the Log Volume and transferring the backup/ log file to the Secondary System. On the Secondary System, the log files are either applied immediately, or queued for a few hours before being rolled into the database.
The status of a single or multiple Standby Databases can be monitored any time using the Libelle GUI. Via the built-in ‘Alarm’ interface, the server agents are also able to send emails, write Windows Event logs or send SMTP traps in case of situations deemed as Warning (e.g. minor network disturbances, timeouts, file transfer problems, or any other issues which are self-corrected by the software) or Errors (Server down, Network down, Database not accessible or break in the log chain, etc.). All messages are stored on both servers and also viewable through the GUI.
Additionally, the TRSAP ‘Control’ interface allows a variety of management functions for the administrator such as starting or stopping processes, edit configuration and parameter files, or restart server agents:
Finally, the last, but most important feature is the ‘Switch-Over’ Interface which manages the startup of the Standby Database on the Secondary System with various options. Failover options include:
(1) Point-in-Time Recovery
Recover to a specific time stamp. Possible if a queue / time-delay is configured to apply logs only to a certain time-stamp and cut off transactions after. This can be used in case of Software-/ User Errors and any form of data corruption on the Primary System.
(2) To End of Logs
Recover all logs which are physically on the Secondary System.
(3) Lossless Switch
Recover all logs on the Secondary System plus try to access the Online Log from the Primary System if still available.
(4) Admin Switch / Defined Switch
During a planned downtime with access to both databases, the Defined Switch will change roles between Primary and Secondary System.
A couple of features are very unique and add much value to customer's operation of MaxDB standby databases. Some of those features include:
- Defined Switch: Libelle supports a 'role-switch' between Primary and Secondary Systems. During a planned downtime, DBShadow server agents are switching roles between Production and Standby System. This allows to temporarily or permanently running production on the Secondary System without recopying the complete database. Using the Defined Switch feature can help in scenarios such as system maintenance in HA-Settings or for area evacuation.
- Recreate Bad-Indexes: After fail-over to the Secondary System, Libelle is able to rebuild all bad indexes automatically.
- Initial Copy via Pipe: To build a mirror between Source and Target System it is necessary to initially copy the database to the target system. We use an automated method for building that mirror called “Initial Copy”. This procedure is not automating and compressing data, it allows to pipe the required backup of the production database directly to the Target system without having to recopy the database
- Alternate Archive Directories: DBShadow/MaxDB is now supporting natively Alternate Archive Directories. This means that backup of log volumes are executed as usual, but copied to a second directory. One directory is now used for e.g. tape backup, the second directory for Libelle to keep replication to Secondary Systems active. Secondary Archive Directories are also usually managed and cleaned by Libelle Software.
- Fast Recovery: The MaxDB database on the Secondary System is in constant recovery mode applying log files. It is usually necessary to issue 'Recover Cancel' commands for logs to be fully applied to the database. While previous Libelle versions did this one log at a time, newer versions allow batches of logs to be applied and thus dramatically improve Recover Performance.
- Dynamic Parameters Changes: DBShadow is designed to have pre-configured configurations for the target databases. Upon failover, the target system can automatically run with different memory and storage parameters if e.g. the target system is sized differently. Additionally, certain parameter changes in the production system are also updated to the target system in a similar manner automatically.
Please contact Libelle to setup a live demo of the solution. For North and South America, please reach our Atlanta Office at [+1] (770) 435 1101 or email@example.com.
For Europe, Middle East, Africa, and Asia, please reach out at [+49] (711) 78335-0 or firstname.lastname@example.org.