Part 5. Recovery and problem determination

Partial Table-of-Contents

  • Chapter 14. Recovery and restart
  • Making sure that messages are not lost (logging)
  • What logs look like
  • The log control file
  • Types of logging
  • Circular logging
  • Linear logging
  • Using checkpointing to ensure complete recovery
  • Checkpointing with long-running transactions
  • Calculating the size of the log
  • Managing logs
  • What happens when a disk gets full
  • Managing log files
  • Log file location
  • Using the log for recovery
  • Recovering from power loss or communications failures
  • Recovering damaged objects
  • Media recovery
  • Recovering media images
  • Recovering damaged objects during start up
  • Recovering damaged objects at other times
  • Protecting WebSphere MQ log files
  • Backing up and restoring WebSphere MQ
  • Backing up WebSphere MQ
  • Restoring WebSphere MQ
  • Recovery scenarios
  • Disk drive failures
  • Damaged queue manager object
  • Damaged single object
  • Automatic media recovery failure
  • Dumping the contents of the log using the dmpmqlog command

  • Chapter 15. Problem determination
  • Preliminary checks
  • Has WebSphere MQ run successfully before?
  • Are there any error messages?
  • Are there any return codes explaining the problem?
  • Can you reproduce the problem?
  • Have any changes been made since the last successful run?
  • Has the application run successfully before?
  • If the application has not run successfully before
  • Common programming errors
  • Problems with commands
  • Does the problem affect specific parts of the network?
  • Does the problem occur at specific times of the day?
  • Is the problem intermittent?
  • Have you applied any service updates?
  • Looking at problems in more detail
  • Have you obtained incorrect output?
  • Messages that do not appear on the queue
  • Messages that contain unexpected or corrupted information
  • Problems with incorrect output when using distributed queues
  • Have you failed to receive a response from a PCF command?
  • Are some of your queues failing?
  • Does the problem affect only remote queues?
  • Is your application or system running slowly?
  • Tuning performance for nonpersistent messages on AIX
  • Application design considerations
  • Effect of message length
  • Effect of message persistence
  • Searching for a particular message
  • Queues that contain messages of different lengths
  • Frequency of syncpoints
  • Use of the MQPUT1 call
  • Number of threads in use
  • Error logs
  • Log files
  • Early errors
  • Ignoring error codes under Windows systems
  • Operator messages
  • Dead-letter queues
  • Configuration files and problem determination
  • Tracing
  • Tracing WebSphere MQ for Windows
  • Selective component tracing on WebSphere MQ for Windows
  • Trace files
  • An example of WebSphere MQ for Windows trace data
  • Tracing WebSphere MQ for AIX
  • Selective component tracing on WebSphere MQ for AIX
  • An example of WebSphere MQ for AIX trace data
  • Tracing WebSphere MQ for HP-UX, WebSphere MQ for Solaris, and WebSphere MQ for Linux for Intel and Linux for zSeries
  • Selective component tracing on WebSphere MQ for HP-UX, WebSphere MQ for Solaris, and WebSphere MQ for Linux for Intel and Linux for zSeries
  • Example trace data for WebSphere MQ for HP-UX, WebSphere MQ for Solaris, and WebSphere MQ for Linux for Intel and Linux for zSeries
  • Trace files
  • First-failure support technology (FFST)
  • FFST: WebSphere MQ for Windows
  • FFST: WebSphere MQ for UNIX systems
  • Problem determination with clients
  • Terminating clients
  • Error messages with clients
  • UNIX systems clients
  • DOS and Windows clients


    • © IBM Corporation 1994, 2002. All Rights Reserved