Maintaining the transaction log file

The transaction log file records the transactions performed on the database. You must delete the transaction pronto.log file when the size of the file increases and causes the archives to fail or the disks to be full.

The size of the log file varies depending on the time the last successful database archive (dbarchive) was performed and the volume of activity in the database. For example, the activity in the database increases in relation to the number of monitors you are using. 

The same transactions are updated in the main database file during checkpoint. The SAP SQL Server automatic checkpoint mechanism writes all pages that have been modified in memory, but not on disk, since the last checkpoint, to the database device. This automatic checkpoint mechanism guarantees that data pages changed by completed transactions are regularly written from the cache in memory to the database device. However, the transaction log file can have some transactions that have occurred since the last checkpoint. The size of the transaction log file can continue to increase until truncated by a full archive job.

You can verify the success and time of the last checkpoint by viewing the database error log. The database error log is usually named storm_<hostname>.log, and it is stored in the following location: 

installationDirectory\pw\asatmp

To verify the success and the time the last checkpoint occurred, you can check the time stamp for the main database error log file storm_<hostname>.db in the pw\sybase directory. The following is an example of the entries made in the database error log during a checkpoint:

When the size of the pronto.log is very large or when the file is corrupted, it can take more time to start the database (as indicated by the error log in the pw\asatmp directory). In this scenario, you must delete the pronto.log file and start with a fresh version of the file.

To delete the pronto.log file

1. Ensure that the database dbsrv process is stopped. If it is running, stop the service by using the pw system stop command. If the process is running or not, can be verified from the output of the pw system status command. The entry for dbsrv appears as NOT Running.
2. Perform the following tasks to rename the pronto.log file in the installationDirectory/pw/sybase directory:

   1. Run the following command to change the directory to the <installationDirectory>/pw/sybase directory:

      (Windows) cd installationDirectory\pw\sybase
      (Linux) cd installationDirectory/pw/sybase

   2. Run one of the following commands to rename and move the transaction pronto.log file:
      (Windows) move pronto.log pronto.log_orig
      (Linux) mv pronto.log pronto.log_orig

3. From the command prompt, run the command specific to your operating system to recover the database file:

   Note on dbeng command

   Use dbeng16 if you are using 11.3.01 or 11.3.02 version.

   (Windows) dbeng17 -f storm_<hostname>.db

   A message confirms that recovery is complete.
   (Linux) installationDirectory/pw/sybase/asa/bin/dbeng17 -f storm_<hostname>.db

   When prompted, enter q to shut down the database immediately.

To create a pronto.log file

To create a pronto.log file, perform the following procedure:

1. Follow the instructions specific to your operating system:

   1. (Windows) From the command prompt, run the command dbeng17 storm_ <hostname>.db

      A message confirms that the request is accepted.

   2. Press Ctrl + Alt + Delete to open the Windows Task Manager.
   3. In Processes, select dbeng17 and click End Process.

1. 
   
   1. (Linux) From the command prompt, run following command: installationDirectory/pw/sybase/asa/bin/dbeng17 storm_<hostname>.db
   2. When prompted, enter q to shut down the database immediately.
      A new pronto.log is created in the sybase directory.
2. From the command prompt, run the pw system start command to start Infrastructure Management.