“A cell is a collection of components, grouped from design and implementation into deployment. A cell is independently deployable, manageable, and observable.”
“In my time as a manger, I’ve encountered quite a few people who were badly onboarded. As a rule it takes as long a time or longer to fix it as it took to poorly onboard them.”
“To get our work life balance back and help the business expand, we took a month to create an automatic remediation system. In this post, we hope to share what we built and what we learned.”
RT @crayzeigh: Human error is never the cause of an incident. It’s an excuse to avoid learning. https://twitter.com/mattstratton/status/1179422688382767105
RT @kstewart: An engineering manager with 60 direct reports is not doing engineering management. https://twitter.com/mattetti/status/1179424874319626241
RT @vllry: “I got myself a gym membership, but if I didn’t do the work, I’m not going to get any better.” https://twitter.com/mweagle/status/1179563249051693056
“Think about incorporating the capabilities the world now offers and removing custom components and infrastructure that are no longer unique to your business.”
(via @Pocket) https://blog.atomist.com/modern-software-world/
“Bigslice provides a coherent set of operators that helps the user efficiently compute over large data sets using ordinary Go code.”
(via @Pocket) https://medium.com/grail-eng/bigslice-a-cluster-computing-system-for-go-7e03acd2419b
Lesson: Crayons made Google successful. We must also use crayons.
Narrator: That is not the lesson.
“This Is Going to Be Huge”: Google Founders Collection Comes to Computer History Museum https://hackernoon.com/this-is-going-to-be-huge-google-founders-collection-comes-to-chm-27356256731f
“I started by noting that solitude and leadership would seem to be contradictory things. But it seems to me that solitude is the very essence of leadership.”(via @Pocket) #longreads https://pocket.co/xd6AzD
“Having a culture where people are not optimizing to avoid mentioning something true but irrelevant means we make better technical decisions overall because we get the benefit of every teammate’s complete perspective.”(via @Pocket) #longreads https://pocket.co/xd6A56
“Some unique characteristics of the npm ecosystem vs other package ecosystems are the high number of transitive dependencies, and heavy reliance on micropackages consisting of only a few lines of code...So what could possibly go wrong?”
Yep. Almost as if when the language change, but the undesirable outcomes remain, and the shop never seems to evolve either way, then perhaps it’s not the technology that’s the limiting factor.
“[W]e just witnessed how a container has been able to affect another container in a non-obvious way, even if they weren’t really competing over any physical resource.”(via @Pocket) #longreads https://pocket.co/xd9WTQ
“Barbara Fusinska from Google demoed the Kubeflow project—designed to automate the deployment of TensorFlow-based models on Kubernetes.”(via @Pocket) https://pocket.co/xdfabi
“I got the fantastic opportunity to moderate a Q + A Session with Scott Guthrie, the Executive Vice President of the Cloud and AI organization at Microsoft.”(via @Pocket) https://peopleofcolorintech.com/articles/cheers-to-year-one-at-microsoft/
“[P]eople are building their entire career around Kubernetes. In a few years, no one is going to care about it at all. It’s going to have simplified and slipped below the surface of awareness.”
RT @sapessi: You can read Raghav’s story here https://link.medium.com/Fb3CIThkm0 and help by donating on this GoFundMe https://www.gofundme.com/f/supportRaghavsresearch
RT @sapessi: The energy, dedication, and positivity of @sanathkr_ and his wife are a true inspiration. https://twitter.com/sanathkr_/status/1183884912291532800
“While it is tempting to make fun of this approach in serverless conferences and meetups, how about we ask ourselves a simple question — what properties can we borrow from this approach?”(via @Pocket) #longreads https://pocket.co/xd1sjn
“Dav Tagare from Lyft described their streaming platforms (160 billion writes / day over 11 clusters) — wow!”
Kafka Summit 2019 San Francisco: a review(via @Pocket) https://pocket.co/xd1sJX
“The lesson from nature and the internet is that the sub-teams should look like a single, small software organisations. How small? Ideally one to five individuals.” (via @Pocket) #longreads https://pocket.co/sd1sqy
“What if you want to trace what happens inside a system call or library call? What if you want to do more than just logging calls, e.g. you want to compile statistics on certain behavior?”
Full-system dynamic tracing on Linux using eBPF and bpftrace https://www.joyfulbikeshedding.com/blog/2019-01-31-full-system-dynamic-tracing-on-linux-using-ebpf-and-bpftrace.html
“I got a bit distracted by the question of whether one could apply SRE procedures to something that does not involve any actual software engineering.” (via @Pocket) #longreads https://pocket.co/xd1q1M
“Slack is not air traffic control that coordinates everything. It’s 911 for when everything falls apart.” (via @Pocket) #longreads https://pocket.co/xd1qZu
“Expressing available SLOs in terms of ‘nines’ also causes some issues: it hides the difference between many short outages and a few long ones, which is something many customers care about.” (via @Pocket) #longreads https://pocket.co/xd1qyU
“In short, the core interface developers have used to compose software for the past 50 years—functions that exchange data through memory—misuses the most precious resources on modern hardware.”
“Machine learning will play an important role in optimizing execution”
https://docs.google.com/presentation/d/1k2rsaMGdsvWrV1HhzDOrVPvq6kDgQj6_0MAZ7NoFl58/mobilepresent?slide=id.g40ccf77c14217d7c_533
There are a lot of good points in this conversation and I would also add that everyone across the organization is typically experimenting and learning. Even if that experimentation is “hidden” behind an API.
“Dapr exposes its APIs as a sidecar architecture, either as a container or as a process, not requiring the application code to include any Dapr runtime code.”
🤔
“When it comes to application development and deployment, we think it is important to distinguish between the parts that developers are responsible for, and the parts that operations is responsible for.”
The Open Application Model specification https://github.com/oam-dev/spec/
1. The developer creates a web application;
2. The application operator deploys instances of that application, and configures it with operational traits, such as autoscaling;
“Moving from using ML and AI for experimental proof-of-concept applications to production systems requires important upfront work and ongoing maintenance.”
“Conferences like this help, but more and better documentation, more sharing of best practices, and tools that can truly streamline the job of delivering business value on top of Serverless remain a work in progress.”(via @Pocket) https://pocket.co/xdaskG
“Serverless functions are fundamentally event-driven. There are two basic ways to trigger a serverless function: directly with an invocation, or indirectly based on something that occurred in the world.”(via @Pocket) #longreads https://pocket.co/xdazP3
“Every layer of the technical stack from development to production has become more complicated, involving more moving parts. What’s unclear is how sustainable this trend is.”
Integrated Innovation and the Rise of Complexity – tecosystems https://redmonk.com/sogrady/2019/10/18/complexity-integrated-innovation/
In Lambda it includes cross cutting information (eg: Deadline()) and props: https://docs.aws.amazon.com/en_pv/lambda/latest/dg/go-programming-model-context.html
Sparta stuffs a logger, session info in there: https://github.com/mweagle/Sparta/blob/master/sparta_constants.go#L82
“This post will serve as a simple end-to-end example of how to use your own tensorflow-model to do inference in your go-application.” (via @Pocket) https://pocket.co/xdeazt
‘I believe we’ll start to see more “architecture as code” tooling emerge that enables developers to describe their microservices in terms of high-level infrastructure patterns.’ (via @Pocket) #longreads https://pocket.co/xdeaje
Gotta make a move to a cloud that’s right for me
Town to keep me devn'
Keep me deploy’n with some energy
Well, I talk about it, talk about it
Talk about it, talk about it
Talk about, talk about
Talk about movin'
...
...
Won’t you take me to
Multicloud?
“How was I supposed to ever learn what that team members was thinking and feeling about the company if I never talked with them?”(via @Pocket) #longreads https://pocket.co/xderuY
“This means ingesting a lot of data, sometimes billions of records per day, which we do in realtime using Kinesis and Lambda.”(via @Pocket) https://read.iopipe.com/lessons-from-building-a-serverless-data-pipeline-with-aws-kinesis-and-lambda-4d8cf0ebcbc9
“Delta is an eventual consistent, event driven, data synchronization and enrichment platform.”(via @Pocket) #longreads https://medium.com/netflix-techblog/delta-a-data-synchronization-and-enrichment-platform-e82c36a79aee
“Serverless computing is one technology that helps Amazon innovate faster than before. It frees up time for internal teams by eliminating the need to provision, scale, and patch infrastructure.”(via @Pocket) #longreads https://aws.amazon.com/blogs/opensource/real-world-serverless-application/
“KSQL enables data scientists to take a look at Kafka event streams and implement continuous stream processing from their well-known and loved Python environments like Jupyter...”
Machine Learning with Python, Jupyter, KSQL and TensorFlow | Confluent https://www.confluent.io/blog/machine-learning-with-python-jupyter-ksql-tensorflow
“After designing and learning a ML model, the hardest part is actually running and maintaining it in production.”
Pure serverless machine learning inference with AWS Lambda and Layers https://medium.com/merapar/pure-serverless-machine-learning-inference-with-aws-lambda-and-layers-979702d9ae49
“Managers have direct influence over the tenure of employees but also have influence over how fast teams can establish norms to work better together.” (via @Pocket) #longreads https://www.fellow.app/blog/2019/10x-managers-supermanagers/
“In reaching v1.0 and moving to Incubation, the spec defines the common attributes of an event that facilitate interoperability, as well as how those attributes are transported from producer to consumer via some popular protocols.”
https://www.cncf.io/announcement/2019/10/28/serverless-specification-cloudevents-reaches-version-1-0/
“What we need is a quick and easy way to go from our big data model to a micro-service that provides just one prediction at a time, on demand.”(via @Pocket) #longreads https://towardsdatascience.com/from-big-data-to-micro-services-how-to-serve-spark-trained-models-through-aws-lambdas-ebe129f4849c
It seems reasonable, but IMO the real power is leaning into a single ecosystem. A common event format will limit my ability to do that/push complexity into the ‘data’ handler.
The Advantage: Why Organizational Health Trumps Everything Else in Business https://learning.oreilly.com/library/view/the-advantage-why/9780470941522/ffirs01.xhtml via @safari
“TensorFlow Enterprise offers a white-glove service to help these cutting-edge customers tackle their biggest AI challenges. It includes engineer-to-engineer assistance from both Google Cloud and TensorFlow teams at Google.”
