Logging best practices to get the most out of application-level logging — Slides Logs from the application level are the only way your application talks back to you. Having proper logs with commendable search capabilities is like having a torch in a dark room. Fully exploiting logs in case of any issues is very underrated. We as software engineers can utilize logs from the application to resolve issues as well as know the general health of the application. Logs can be instrumental to profile application performance too. I had written about some years back too. This post is focused on best practices you can follow when logging from the application level and how it can help you as a software engineer. logging Why log from the application level Logging Best practices at a glace Following are the main best practices for logging: 1. Log information optimally Too much information is noise and too less information is inadequate. It is difficult to strike the balance of how much log is optimal but that is the challenge. In the case of microservices also think about traceability between services like using a unique request identifier. Another thing to keep in mind is logs are temporal, not permanent. As they are not stored in the database they usually have a life from some days to some weeks. 2. Always follow severity levels An emergency means your on-call phone is going to ring at 2 AM in the morning but an info severity is something that won’t bother anyone. You have to agree as a team on some set standards like the . syslog one 3. Structure your logs Have a structure in your log lines, like have a message and a context array to add more information about the message. Follow an agreed upon JSON standard for logging. This makes it easy to parse and search. You could even set more rules like the date is mandatory, a description cannot be more than 255 characters, extra information goes in the context. Having these simple rules makes it easier to find the log when in need. It is highly recommended to provide context with log lines like add the details of the order (things that are not sensitive of course) when logging that the order could not be shipped for instance. 4. Write logs carefully (don’t hamper performance) Make sure that adding logs does not add a lot of extra milliseconds to your response times. As far as possible write logs in an async way, logs could even be written locally in a log file and shipped using a log shipper to the service of your choice. You can also make use of tried and trusted library depending on the language/framework you are working with. Using queues could also be an option to ship logs but keep in mind the possible lag when viewing logs. 5. Use the right tools for the job The tools and services you choose to ship, view, search and sort logs are your team’s choice. Depending on the budget, the appetite for using external services a team could use a fully self-managed stack like / or go full SAAS with services like / . The main idea is to use the tools that fit your needs very efficiently. If you need almost real-time logs using a queue, for instance, might not be the best decision. So design your logging infrastructure and stack that fits your needs well. ELK Graylog Logentries Sematext Logsense The slides I had done a talk at my workplace in February about “Logging best practices” and the slides from the talk are below: <a href="https://medium.com/media/00980b0e85eb2a452f51d1017c59342a/href">https://medium.com/media/00980b0e85eb2a452f51d1017c59342a/href</a> You can view the slides on and too. slideshare speaker deck Conclusion If a dog is a man’s best friend, logs are software engineer’s best friend. Use logs to the fullest to get the most out of it and keep in mind the best practices about logging. Originally published at geshan.com.np .

Keep

Mind

Tail logs from multiple Kubernetes pods the simple way

What Tech Recruiters Actually Look for in your Resume and First Interview

Follow these logging best practices to get the most out of application-level logging — Slides

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

Untitled Story

3 Non-Technical Qualities of a Great Software Engineer

10 Things I Did To Increase CloudTrail Logs Security

44 Stories To Learn About Logging

5 Tips for Integrating Security into Development: Part 2

5 Tips to Effectively Monitor Heroku Applications

3 Non-Technical Qualities of a Great Software Engineer

10 Things I Did To Increase CloudTrail Logs Security

44 Stories To Learn About Logging

5 Tips for Integrating Security into Development: Part 2

5 Tips to Effectively Monitor Heroku Applications

Light-Mode

Classic

Newspaper

Dark-Mode

Neon Noir

Minty

HN StartUps