IBM i system monitoring software watches an IBM i partition 24/7/365 and immediately takes action after detecting hardware, operating system, or software issues. When finding a problem, the software can alert staff via email or text messages, or it can initiate action on its own to resolve an issue.
The most common system monitoring technique is to monitor messages in the QSYSOPR message queue or the QHST log, and then send out alerts based on message content.
If you’re evaluating IBM i system monitoring software or if you’re starting to configure a monitoring package, here are six message monitoring techniques you can use to get the most out of your software:
1. Monitor for Messages with Individual or Generic Message IDs
Every system and application-generated message has its own seven character message ID. Once you know the message ID that you want to monitor for, you can configure a monitor to search for messages with that ID in a specific message queue (i.e. monitor for message ID RPG1216 in the QSYSOPR message queue). The software would then send out an alert every time that message appears in the designated message queue.
You can also monitor messages by generic message IDs. Simply enter any part of the seven character message ID followed by an asterix (*) into a system monitor. For example, if you wanted to monitor for all RPG messages in the QSYSOPR message queue, you could set up a monitor to watch for message ID RPG* in QSYSOPR. The software would then take action every time it saw a generic RPG* message posted to QSYSOPR (RPG1216, RPG1217, etc.).
2. Monitor for Specific Text Inside a Message
Some monitoring packages including our own MessageFlex software, can screen messages for specific text values. Message text scanning is valuable when sending alerts for one specific instance of a posted message, instead of sending alerts for many system messages with the same message ID.
3. Monitor for Message Severity
All messages contain a severity code from ‘0’ to ‘99’ that specifies how critical the message is. The higher the severity code, the more critical the message. With most packages, you can set up monitors to trigger alerts when a severity ‘99’ message is received or for any messages that exceed a certain severity level (such as ‘50’). This technique is used to send alerts for all urgent messages in a message queue.
4. Monitor for Specific Message Types
The IBM i operating system divides messages into different message types according to their usage. Each posted message contains a message type. System monitors can be configured to trigger alerts after finding a new message containing a specific message type, including:
- Escape (*ESCAPE) – A program ended abnormally, without completing its work
- Inquiry (*INQ) – A message that’s waiting for a reply, such as a program error, file full, unable to allocate record, etc. Tip – monitor most *INQ messages relating to program activity but skip *INQ messages for common items like printer alignment or form change messages.
For a complete list of message types, see IBM’s Message Handling and Terms Web page.
5. Combine Message Monitoring Techniques
Most system monitoring packages allow you to combine any or all of these techniques into one monitor. For example, you could set up a monitor to look for all RPG* messages in QSYSOPR that are inquiry messages (message type *INQ). Or you could set up a monitor to look at all messages with severity 50 and above that contain specific text. Most monitors can also be set up to look for messages generated under certain job names, user profiles or in specific subsystems.
There are several different ways you can mix and match message monitoring techniques to detect the information you need, and nothing else.
6. Exclude Messages from Monitoring
There will always be IBM i messages that don’t require action, even if they trip a system alert. One example would be printer alignment inquiry messages which shouldn’t be forwarded to your incident responders.
Most system monitoring packages can add ‘ignores‘ to a monitor. An ‘ignore‘ tells the monitor to skip sending an alert if the monitored message comes from a certain job, user, subsystem, or other criteria. ‘Ignores‘ reduce the noise in your system monitoring environment and help make sure you only receive the alerts that you really need.
Understanding how message monitoring works is a great way to get started with an IBM i system monitoring package. Please feel free to contact us at DRV Tech if you need help evaluating or getting started with your own system monitoring solution. And don’t forget to check out the helpful articles related to this post below for more system monitoring strategies, tips, and techniques.
Want to learn more about MessageFlex? Click below to see MessageFlex Fast Facts.
Interested in seeing MessageFlex in action? Schedule a free demo with DRVTech by clicking below.