APMdigest asked experts from across the IT industry for their opinions on what IT departments should be monitoring to ensure digital performance. Part 2 covers key performance metrics like availability and response time.
Start with What You Should Be Monitoring to Ensure Digital Performance - Part 1
AVAILABILITY
To ensure digital performance, availability is one of three key performance areas I always recommend monitoring. Your applications and networks must first be available to service users and customers. Otherwise, they're useful to no one.
Jean Tunis
Senior Consultant and Founder of RootPerformance
Monitoring the login page of an application with a synthetic transaction is an essential part of an Enterprise Monitoring strategy. Active monitoring is a good starting point to provide visibility on application availability especially when monitoring outside the Data Center. Synthetic Transactions can provide location-based availability and act as a barometer for measuring application performance.
Larry Dragich
Technology Executive and Founder of the APM Strategies Group on LinkedIn.
Read Larry Dragich's latest blog: Digital Intelligence - Why Traditional APM Tools Aren't Sufficient
Read Larry Dragich's new white paper: The Case for Converged Application & Infrastructure Performance Monitoring
INFRASTRUCTURE RISK
Understanding Infrastructure Risk is a key component of monitoring that most organizations miss. APM tools do a great job of tracking left-to-right performance across an application, and modern application designs ensure that no single component can cause a failure. Building an understanding of the risk inherent in the current IT infrastructure — below the application — is critical for stopping unexpected downtime and sudden capacity limits. You can do that by tracking links from between overlay and underlay networks, file systems to storage units, and hypervisor to server hardware — or you can use a unified monitoring tool do do it for you. Key buying decision — can you see the IT infrastructure risk for the specific components that your application relies on?
Kent Erickson
Alliance Strategist, Zenoss
THROUGHPUT
To ensure digital performance, throughput is one of three key performance areas that must be included. Applications and networks must be able to provide all the relevant data that is required to fulfill a specific request. Monitoring throughput ensures you know when your systems do not deliver all of the data that was requested, and you can act on it before the complaints come in.
Jean Tunis
Senior Consultant and Founder of RootPerformance
PAGE LOAD SPEED
Ultimately you want to be monitoring everything that impacts customer experience and conversion rates, but the most important thing is the page load speed. This drives more conversions than any other factor. And the key pages are those at the beginning of a user journey since the more time someone has invested in the process the less likely they are to abandon.
Antony Edwards
CTO, Eggplant
RESPONSE TIME
To ensure digital performance, response time is one of three key performance areas that must not be forgotten. Requests for specific information from users must be fulfilled with as much speed as possible. This is a common expectation of every IT system, so you should be monitoring them.
Jean Tunis
Senior Consultant and Founder of RootPerformance
Monitor application response from user to application (last mile) and application to the data (middle mile) to not only measure is the app up but of it is working.
Jeanne Morain
Author and Strategist, iSpeak Cloud
TRANSACTION UPTIME
A good starting point is to implement end-to-end performance monitoring with real transaction uptime to complement your APM tools.
Sven Hammar
Founder and CSO, Apica
TIME TO FIRST BYTE
Initial motivation in the user journey can be lost very quickly if for example the first time the user clicks on an advertisement or logs into an application is not performant. The appearance of performance is important; monitoring time to first byte (TTFB)can help ascertain the experience of what a user sees marching towards a minimum viable/viewable product (MVP) of the page or app before being loaded to completion. TTFB is a leading indicator on web performance to the end user and also is used by the leading search engines in factoring in page rank as the more performant pages get a higher rank.
Ravi Lachhman
Evangelist, AppDynamics
LOG EVENTS
If it has an IP address, it sends logs, and logs must be monitored to gain detailed insight on server performance, security, error messages or underlying issues.
Clayton Dukes
CEO, LogZilla
Logs have been around since the dawn of computing, but with constantly increasing threats, logs are more important than ever. Log events are one of the key data sources SIEM (Security Information and Event Management) solutions use for threat detection.
Otis Gospodnetić
Founder, Sematext
Read What You Should Be Monitoring to Ensure Digital Performance - Part 3, covering the development side.