Message Recovery Recommendations

  • Ensure sufficient disk space is allocated to the local file system so it does not run out of space. Disk space requirement can be calculated based on the total number of expected messages in a system failure. For example, for 10K messages and 1000 messages/minute, the system failure for about 10 minutes would need about 10 MB of disk space. It is recommended you allocate space equivalent to at least three times the maximum expected message load.
  • In a clustered environment, each application server generates one messages-redo.log file. If this file is configured to be written to local disk, in case of an error, the failed incoming JMS message can be in any one of the application servers configured target directory. To recover all failed messages, you need to run the message recovery tool on each server. If there are any dependencies between failed messages written in different servers, manual consolidation of the messages-redo.log file and message’s serialized object files from each server may be needed.