5 tips for setting up successful and dependable IT infrastructure monitoring
July 27, 2019
Here are my top five tips for setting up successful and dependable IT infrastructure monitoring:
5. Be forewarned!
We know monitoring can’t tell the future... but we also know there’s tech that at some point WILL fail. An obvious example is hard drives: they don’t usually fill up “all of a sudden”, and even full-on failures usually come with warning signs. Also think of room temperature, which tends to increase slowly in most cases. You need to be forewarned at the right time to effectively handle these predictable events.
4. Supervise your backups automatically
An often forgotten but crucial step in effective monitoring is keeping an AUTOMATIC eye on your backups. Maybe you get emails when backups run and fail… but you shouldn't rely on these to warn you of massive failures of your backup systems. Because if it crashes completely, no email is sent. And your backups are your endgame! They are your last line of defense in recovering from a crashed system. Ensure you have a regular and automated process - running at least daily - to check that a functional and recent backup of your data exists.
3. Monitor all your single points of failure
Every systems setup has critical points of failure, components which can “turn off all the lights”. Think about your routers and switches. If they fail - it can ripple across your entire setup. You need an immediate alert when those devices fail, even if they are not under your direct control. It’ll help you to predict and react in advance to other alerts which come along for the ride with a core-router outage, for example.
2. Reduce false alerts
Effective and dependable monitoring is an ever evolving process. You should continuously remove any checks you don’t need to cut down useless and distracting alerts. Be sure to only get the most critical alerts on your mobile… where critical means you MUST take action. I’ve found that just a few false alerts are enough to lose trust in your monitoring solution. If an alert comes in and you aren't sure how or if to react, or even if it's real… - you‘re lost. Your monitoring has converted to a spam cesspool, and you’ll begin to ignore the alerts and warnings, making is basically useless! A strict differentiation between alerts and warnings is also helpful.
1. Start monitoring from your customers’ perspective
The first and most important assets you need to monitor are the services your customers are using. Monitor what your customers pay for. Typically these are things like TCP ports responsible for essential network services, like port 80 for your web server, port 143 for your imap server, and so on. If these ports don’t answer correctly, you can assume your customer is suffering an outage of some sort. Monitoring network ports is simple and easily overlooked, but so crucial.
CloudRadar provides Server, Network and Website Monitoring with a focus on fast configuration, ease-of-use and affordability. You can monitor your entire systems infrastructure, so you have all your systems under control from one central modern cloud monitoring solution..
Start your Free Trial of CloudRadar now >
Free 15 Day Trial. No credit card required. Then choose from our simple pricing plans.
Author: Susanne Weiss/Nicholas Thiede