Interstage Shunsaku Data Manager Troubleshooting - Microsoft(R) Windows(R) 2000/ Microsoft(R) Windows Server(TM) 2003 - - UNIX -

2.3.4 Managing Disk Capacity

This section explains how to avoid overloading the Shunsaku disk resources.

• Managing the Director Data File.
• Managing the Operating Log File.

Managing the Director Data File

To avoid overloading the capacity of the director data file, manage the file as follows:

• Managing the director data file storage directory:

  The following procedures can be used to manage the director data file storage directory:

  • Estimate the size of director data file.
  • Estimate the size of director data file.
  • Estimate the maximum size and the alarm point of the director data file.
  • Estimate the maximum size and the alarm point of the director data file from the estimation result of the director data file size. Make sure the size of the director data file storage destination is equal or more than the maximum size.
  • Only store the director file in this directory. Do not store any other files in the same directory.
• Managing operations:

  Deleting or updating large amounts of search data may increase the amount of wasted space in the director data files and search performance may drop. The more wasted space the less space is available for the director data file storage directory. It may not be possible to optimize the director data file by executing the shundcds command.

  In this case, the following procedures can be used to avoid the above problems:

  • Use the following parameters in the director environment file to set the maximum size and the alarm point for the director data file:
    • Maximum size: MaxDataFileSize parameter.

      If the data file size exceeds the maximum size, updates will not be processed and all the transactions, including the update process, will be rolled back.

    • Alarm point: WarningDataFileSize parameter.

      If the data file size exceeds the alarm point, an error message (shn30395w) is output to the system log (syslog) or event log.

  • Monitor and optimize the director data file:

    Use the monitoring function (shundstate command) to monitor the total amount of director data and the fragmentation rate. Also, optimize the director data file by periodically executing the shundcds command.

Refer to Execution parameters in director environment files and Monitoring logging in the Operator's Guide for more information.

Managing the Operating Log File

To avoid overloading the capacity of the operation log file, manage the file is as follows:

• Manage the operation log file storage directory with the following procedures:
  • Specify a backup cycle for the director data file. Specify the capacity of the disk to allow for the size of updated log data stored in the operation log file. The amount of updated log data is the amount of data updated between backups. Use this amount to estimate the size of the operation log file. Then allocate enough space in the storage directory.
  • Use the estimate of the operation log file size to estimate the maximum size and the alarm point of the operation log file. Allocate enough space on the storage directory to store the maximum size.
• Managing operations:

  Use the following procedures to manage the operation log file:

  • Set the maximum size and the alarm point of the operation log file with the following parameters in the director environment file:
    • Maximum size: MaxOperationLogSize parameter.

      If the file size exceeds the maximum size, updates will not be processed and all the transactions, including the update process, are rolled back.

    • Alarm point: WarningOperationLogSize parameter.

      If the file size exceeds the alarm point, an error message (shn30397w) is output to the system log (syslog) or event log.

  • Monitor the operation log file and change the backup cycle:

    After an operation, use the monitoring function (the shundstate command) to monitor the size of the operation log file and check whether there is sufficient disk space.
    If the size of the updated log data is more than the size specified and the disk is overloaded, execute back ups more often to avoid the error with an insufficient space on operation log file.

  Refer to Execution parameters in director environment files and Monitoring logging in the Operator's Guide for more information.