Lost in the plethora of logging frameworks and online monitoring solutions? This talk will give you the map of the new logging landscape and help you to design or redesign your logging architecture.
If you think logging is boring, stop yawning and think again.
Designed for the ages of monolithic synchronous applications running in data centers or on the desktop, traditional text-file-oriented logging frameworks such as log4net or nlog are becoming obsolete. New tools, best-practices and standards are emerging to make sense out of massively distributed and asynchronous applications that run in the cloud, in containers, or in mobile devices. Solutions for a world where storage and computing power are cheap and can be provisioned within minutes.
Starting with the question “what do we want to log and why?”, this talk gives a bird’s eye overview of the new logging landscape. It introduces open-source solutions like Elasticsearch, Fluentd or OpenTracing, and commercial services like Application Insights. On a code level, it shows how PostSharp can help to generate highly detailed logs with minimal efforts.
The key takeaway of this talk is a broad understanding of the business of logging and application instrumentation, so you’re ready to make choices on your own.
1. Logging in the cloud
a. Logging is changing
b. Expectations of operations vs application development
c. Logging in the cloud: the log processing pipeline
d. The Elastic Stack: example with Serilog
e. A list of open-source and commercial solutions
f. Buyer’s wish list: how to evaluate a solution
2. Automatic code instrumentation
3. Correlating distributed logs
a. the OpenTracing specification
c. DIY log correlation with Serilog: example with ElasticSearch
d. Hierarchical ids
4. Statistical processing: semantic vs structured logging
5. Sampled logging
UPOZORNĚNÍ: Přednáška bude prezentována v anglickém jazyce.
Čas konce přednášky je pouze orientační.