When evaluating IT Infrastructure Monitoring, what aspect do you think is the most important to look for?

1) Ease of deployment and maintenance. The ideal solution will auto-discover your environment and have intelligence built in to tell you what to monitor and how to monitor with built-in alerts that leverage industry best practice thresholds. This way users can anticipate issues and resolve them before users are impacted. 

2) Historical, real-time, and discrete data that will show all IT infrastructure elements used to deliver a single end-user experience. The is the only way to monitor and troubleshoot issues is to have full visibility into the true user experience.

3) Document all user activity, behavior, and system performance so that you can share, integrated, and enhance data to collaborate with management, other IT teams, application vendors, and even end-users. 

I think you need to look for:

1. Unifications. All aspects of the monitoring should be done by one or multiple tools. As an option, integration between tools should be possible.

2. Plug-in based or open architecture. Open Source will be a huge plus. In this case, you will have community support, and hiring the expert for widely used technology should not be the issue.

3. Tools should have quick support - monitoring could go down when you really need this. Open Source tools allow you to have a big market of engineers with good expertise.

4. Agree with other comments - ROI is very important here.

The ability for the solution to correlate data from across the enterprise to remove noise in alerts, and for the alerts to be able to trigger automation to remediate a known problem/incident.

Most monitoring solutions have similar capabilities. Each one has a bit different set-up and points that are stronger than others. According to the 2020 CIO report from Dynatrace, on average, companies get close to 3,000 alerts daily. That is my concern.... Over time, you'll get more than one monitoring system and need to make sure that you have resources to deal with the alerts... scripts with Jenkins will not work as there are so many types of alerts. Try looking at low-code automation that will help you set up instant remediation pipelines. Companies like Lacework created a library of instant remediation workflows that customers can use in minutes...

I would propose to look at Infrastructure monitoring from a different perspective. The corollary I would use is to equate infrastructure monitoring to a big data problem with the need for automation. In today's world we have many infrastructure devices that transmit a large amount of data or telemetry and the key to quick automated response is to look at adjacencies and quickly determine corrective action. I suggest injecting the telemetry into an infrastructure data lake and apply some ML & AI applications to determine issues and automation to quickly solve. The amount of data produced has become daunting and I suggest taking a data driven approach instead of siloed Infrastructure monitoring tools.

Most important: know in depth your environment and the future evolution, upcoming change - this includes systems, solution, operations model, SLO.

After, you can initiate the second step, identify tool(s), etc.  

I would consider multi metrics to evaluate an infrastructure such as:

1- CAPEX, OPEX & ROI for financial.

2- Security, Reliability, operational complexity lifetime ...etc for technical.

3- Extra benefits that might be gained from different solutions such as possible cross solutions integrations which can be a tie breaker in many times.

I hope this helps.

Peace of mind, Easy of use, ROI

It's integration, helping to improve management capability, and ROI, how does it drive value (this include price).

Our most important criteria include price, compatibility with our existing infrastructure, HIPAA compliance, and security.