When you install IBM i system notification and monitoring software, you can eliminate having dedicated staff watching your partitions. With monitoring software, the system alerts you when it sees an issue.
But before you can configure a system monitoring package like MessageFlex, you’ll need to create a monitoring strategy that can be implemented in your software. This strategy defines how on-call responders and your software will react to system issues.
Here are five common strategies for setting up system monitoring for your responders and for your IBM i system monitoring software.
Strategy #1: Only have one person responding at a time
We recommend having only one on-call responder take responsibility for reacting to system alerts and coordinating responses.
You can rotate this responsibility by shift, day of week, weekly, or even monthly. But it’s important to have only one person at a time responsible for answering alerts; otherwise you may create fire drills where multiple responders work against each other trying to solve an issue.
Strategy #2: Set up escalation lists for backup coverage
Determine what to do if the primary responder isn’t available. Mobile devices can run out of power, get turned off, or stop transmitting in low-coverage areas. If the responder can’t receive the alert, they can’t solve the issue.
Most monitoring packages allow you to create escalation lists for your alerts. These lists define how to contact backup responders when the primary responder isn’t available. If the primary responder doesn’t sign on and turn off an alert within a set amount of time, the software will send the alert to the next responder listed in the escalation list. The software will then wait for that responder to sign on and turn off the alert before moving on to the next responder in the list. And so on, until your IBM i monitoring software reaches the end of the escalation list.
The important point is to set your monitors up with escalation lists for backup coverage.
Strategy #3: Create a call tree for issue resolution
Your primary responder won’t be able to handle every IBM i monitoring alert. Many times, they’ll have to hand off an issue to another responder for resolution. That’s where a call tree comes in.
A call tree defines who to contact for resolving different types of issues, including:
- Programming – Program error messages, malfunctioning jobs, other applications issues
- Operations – Machine issues including hardware failures, disk problems, as well as backup issues and job scheduling problems
- Communication failures – Network issues that affect IBM i processing
- Companion server failures – Issues with exchanging data with non-IBM i servers
Create a call tree for your monitoring team so that your primary responders know who to call to resolve different types of issues. And make sure every responder has a copy.
Strategy #4: Give your on-call responders a break
Be flexible when primary responders are handling overnight issues. Don’t expect them to be on time after they’ve handled a 2:00 AM issue. Modify the schedule for chronic issues, where a responder has to handle the same problem several nights in a row, until the fix gets in.
On-call monitoring can lead to burn-out, when a responder is constantly resolving issues day and night. That’s why a sensible rotation schedule for primary responders is important.
Strategy #5: Go slowly when setting up IBM i monitoring
After installing monitoring software, you’ll be tempted to immediately create monitors for every conceivable issue. But new monitors usually need a break-in period. People can make mistakes in monitor set up, which causes unnecessary interruptions and false alarms.
Phase in automated monitoring. Configure your most critical monitoring tasks first. Leave more complicated monitors for later after you’ve settled into your monitoring strategy. If you’re phasing out on-site 24×7 staff, use them as test subjects until you shift monitoring responsibilities to off-site staff.
Strategy & Implementation
It’s important to work out the strategy behind your IBM i and iSeries monitoring as you move towards automated monitoring and a lights out Data Center. Implementing these strategies with IBM i system notification and monitoring software such as MessageFlex, will make for a successful transition.
See how automated IBM i system monitoring can help you create a lights out Data Center by requesting a 15-minute live demo of DRVTech’s MessageFlex software or by downloading a 15-day trial version of MessageFlex.
JAN