Whether you are a veteran kernel developer or someone that is just curious to what their kernel does, ftrace is the tool to use to see what your kernel is doing. Ftrace, is the official tracer tool of the Linux kernel, that does much more than mere tracing. You can graph into the functions that are called within any function in the kernel, live while it is running. You can see what functions call a specific function for a given load. This talk will give you a taste of the depth of what ftrace has available for you.
Haven't we all said at least once, "Burn it with fire! Let's start this whole thing from scratch!" We came pretty darn close to that. Working around the limitations and cruft of a monolith can be frustrating, but is breaking apart an application running in prod worth the risk to uptime and availability? This is the story of how PagerDuty decomposed our core product into distinct services with zero downtime and zero regrets.
Scalability is a hard topic and scale Logging in a distributed environment is not an exception. In this presentation we will dive into the internals of logging for Kubernetes clusters and share best practices for advanced filtering, performance optimizations and suggested mechanisms to deal with backpressure.
Deploying stateful applications on top of Kubernetes is very exciting topic nowadays. In this hands on talk, Alexander Rubin from Percona will talk about how to deploy databases (MySQL, Mongo) on Kubernetes.
As we all know, avocados are full of fat, but if used at the right times, in the right ways, with the right combinations of items, they can be amazingly beneficial. In this lightning talk, I’ll give you 6 ways to use avocados to explain the value that Developer Relations teams bring to the table.
Steps to improve online privacy and security. Concrete technology to keep your personal information from leaking ( being exfiltrated ) from your personal devices.
A non-technical survey of security and privacy tools and settings for people with an average threat model.
* Browser privacy and security addons
* Secure connection tunnelling ( https, SSH, VPN, TOR )
* Email client configuration
* Data liberation
* Data escrow
Problems with single hosts are challenging enough. Scaling up to hundreds or thousands of running hosts only multiplies the problems. However, troubleshooting and remediating production issues at scale can also be much easier to deal with than issues on smaller installations. In this talk, we will explore a real production issue or two around a python application to highlight some sound techniques and approaches to handling services at scale. We will explore some specific howtos from a simple level, such as reading logs, to more complicated things pertaining the overall state of the runtime.
Ever used a database? Wondered where it magically puts all your data? Wondered what "transactions" are and exactly why we need them? Heard some very technical-looking folk talking about "ACID"? Then this is the talk for you.
In this demo-documentary of a single byte, of the life-cycle of a single byte from birth to death, we attempt to demystify the internals of a relational databases' storage mechanism using PostgreSQL as an example.
In some use-cases, Redis can double its memory use during a snapshot operation (replicas attaching, periodically for backups, rewriting AOF, as requested for backups, ...), generally requiring substantial "copy on write" overhead for heavy write loads, up to 100% in some cases. We demonstrate a technology that reduces this memory usage to .3-.5%, or roughly 1/200 the overhead of existing Redis snapshots, while seeing little difference in throughput or latency.
"I'm sure if I work hard, my contribution will one day be recognized and I'll get a raise." --Everyone At Some Point In Our Careers
The world is unfair; this talk won't do much to fix that-- but it will give insight from two industry veterans on how to maneuver within the world we currently inhabit.