Eric Hexter on DevOps Diagnostics - Episode 011

In today’s episode, Jeffrey is joined by Eric Hexter, the Chief Technology Officer of Quarterspot LASO — a fintech company that produces lending platforms using machine learning. He built the company in the Cloud with DevOps as a foundational component of delivering the product. Eric has filled roles as CTO, Chief Architect, Developer, and Consultant. He’s spent most of his career working with web technologies, with a total of twenty years experience producing technology solutions that deliver business value. As Jeffrey says, Eric is the DevOps King. He’s done some incredible work over the past decade and a half and has even written some books — well, namely one book back in 2012, ASP.NET MVC 4 in Action, co-authored by Jeffrey as well. Eric and Jeffrey talk all about DevOps Diagnostics today, running through the various categories within it, such as: system metrics, log files, air conditions, heartbeats, and data integrity checks. Eric also gives his recommendations to those new and experienced with the system — tools, resources, and services. Topics of Discussion: [:47] About today’s guest, Eric Hexter. [1:19] Jeffrey welcomes Eric to the podcast. [2:28] How Eric first got interested in the world of DevOps. [4:01] Eric talks about some of the key points made from his presentation on the Azure DevOps User Group on Meetup. [6:46] What Cloud DevOps diagnostics consist of. [8:10] What categories .NET developers need to be watching in order to operate their systems effectively. [9:08] Eric talks about one of these categories: his favorite system metrics. [12:15] Eric gives a quick rundown on queue-based processing. [14:23] Eric’s favorite queue at the moment. [15:21] The importance of having metrics on every running piece of your application. [18:23] A word from Azure DevOps sponsor: Clear Measure. [18:52] How Eric looks at system metrics. [20:06] On Eric’s team, who looks at the metrics? [20:34] Eric gives an explanation of the next category of Cloud DevOps diagnostics: log files. [22:32] What Eric recommends developers should be logging to a text file. [23:33] Eric explains what a decorator pattern looks like in code. [24:42] Eric briefly explains built-in log files. [25:03] How Eric brings all these log files together to cohesively view them all. [26:31] How does Eric log files? [27:30] Why logging as a first class feature within the application can be incredibly useful. [29:14] The next category of Cloud DevOps diagnostics: air conditions and common patterns within it. [38:13] The next category: heartbeats. [42:00] Eric dives into the next diagnostics category: data integrity checks. [44:04] The differences in structured logging compared to regular logging. [48:46] For structured logging, does Eric have a favorite library? [50:41] Eric’s recommendations tools and services to get started in all of this. [53:06] Eric’s advice on how to consolidate all your flat files to start a consistent view. [54:16] Eric’s favorite heartbeat source. [55:00] Additional resources Eric recommends listeners to go check out after this week’s episode. Mentioned in this Episode: Quarterspot ASP.NET MVC 4 in Action, by Eric Hexter, Jeffrey Palermo, Jimmy Bogard, Matthew Hinze, and Jeremy Skinner Azure DevOps Azure DevOps User Group on Meetup PREMIS Azure Storage New Relic Stackify Application Insights Azure Queue Storage Clear Measure (Sponsor) Grafana Loggly The Netflix Tech Blog Want to Learn More? Visit AzureDevOps.Show for show notes and additional episodes. Follow Up with Our Guest: Eric Hexter’s LinkedIn