The Memory Trap
February 1, 2020
I've been using Twitter for awhile now as a sort of notebook of sorts, capturing all the rando bits of technology news I come across in my daily rounds about the Internet. Epiphanies, observations, or factoids from blogs, podcasts, or video technical sessions are all captured there.
Most of the information would not necessarily be considered news but in each case it is news to me. Each Tweet is a stan-alone idea, or thought. Sometimes, I can put together these random factoids into stories for my day job. Sometimes they are useful for personal projects. Sometimes they just end up as dead ends, albeit interesting ones.
Given how difficult Twitter makes it for me to find my old Tweets later, this column, "Teknik," will serve as a compendium to them. Every so often (monthly? Weekly? Not sure yet), I will compile my recent Tweets into one of these columns, subgrouped into category. I may add additional commentary, if needed.
In January 2020, for instance, I spent a lot of time learning about Kubernetes, the open source container orchestration engine that is paving the way for something called cloud native computing. I also delved into data management, software development, and found a great quote about what artificial intelligence really is.
Every business needs a fact table, an append-only immutable event log of everything that has happened, chronologically ordered & taxonomically uniform, which is easily consumable by a variety of analytic software—@jaykreps explains @ApacheKafka, @seradio https://t.co/CS5QZZRd3b
— Joab Jackson (@Joab_Jackson) January 2, 2020
“Data-driven Infrastructure” could be the natural evolution beyond software-defined networking (#SDN), curating all operational data to improve performance and bring parity with resellers/vendors who rarely promise outcomes — @etherealmind @packetpushers https://t.co/tPIIRkwEa5
— Joab Jackson (@Joab_Jackson) January 8, 2020
Would you pay ~$300/week to learn to be a data analyst? https://t.co/lVGQbopqNp
— Joab Jackson (@Joab_Jackson) January 8, 2020
#NewSQL databases were created to solve "the write scalability challenge of monolithic SQL databases," but they fell from usage as "multi-zone, multi-region and multi-cloud deployments" of distributed SQL databases increasingly became the norm https://t.co/bWShBt74z3 #DataScience
— Joab Jackson (@Joab_Jackson) January 28, 2020
"Experience has taught us that micro-optimisation often has little or no beneficial effect on performance, but does make code harder to understand and maintain." @_pitest https://t.co/v7D3wEVtXK
— Joab Jackson (@Joab_Jackson) January 2, 2020
#Python 3.0 "was released on December 3, 2008. And it took the better part of a decade for the community to embrace it. This should be universally recognized as a failure." https://t.co/0kuJzpqomi
— Joab Jackson (@Joab_Jackson) January 13, 2020
Is debugging a language through #WebAssembly impossible? -- Erik McClure asks.. https://t.co/wtWwZaFG4O
— Joab Jackson (@Joab_Jackson) January 13, 2020
"Sorting is hard, especially when data is distributed." @NicholasStrayer https://t.co/MZ9M3PS17r #DataScience
— Joab Jackson (@Joab_Jackson) January 27, 2020
“Don’t use ZFS. It’s that simple. It was always more of a buzzword than anything else, I feel, and the licensing issues just make it a non-starter for me.” -- Linus Torvalds https://t.co/2VSfdUjXPU by @jlwallen on @thenewstack #OpenSource #Oracle #ZFS #Linux
— Joab Jackson (@Joab_Jackson) January 10, 2020
"Modern Apps are written in Modern Languages, get deployed onto Platforms, and that deployment process is Continuous Delivery (as these things are all interconnected)." @cpswan https://t.co/Wxxy6wEdu7
— Joab Jackson (@Joab_Jackson) January 9, 2020
"Unless you want massively slower build times, larger images, more work, and the potential for obscure bugs, you’ll want to avoid Alpine Linux as a base image" for #Python applications in #docker containers -- @itamarst https://t.co/3UsEMn23ad #AlpineLinux #Linux #GCC
— Joab Jackson (@Joab_Jackson) January 29, 2020
(observation)
Our field isn't quite "artificial intelligence" -- it's "cognitive automation": the encoding and operationalization of human-generated abstractions / behaviors / skills. The "intelligence" label is a category error
— François Chollet (@fchollet) January 7, 2020
https://t.co/66iOINaggY pic.twitter.com/rnVyV1GDTQ
— Joab Jackson (@Joab_Jackson) January 23, 2020
The #Kubernetes Cluster Management API, now in Alpha, is designed to make it easy to provision clusters across different cloud providers (GCP, AWS etc). Also could be good for devs to test on a dedicated cluster — @k_gamanji, Condé Nast, @InfoQ podcast https://t.co/19N0recg7c
— Joab Jackson (@Joab_Jackson) January 6, 2020
While #Istio is touted as a #ServiceMesh for #microservices, the implementation of the Istio control plane will be gradually moving away from a microservice design in favor of a more monolithic design, by necessity -- @christianposta https://t.co/u9oyFj0Dbb #CloudNative
— Joab Jackson (@Joab_Jackson) January 8, 2020
"To keep up with the increased Droplet demand, we were adding more and more servers to handle the traffic. Each new hypervisor meant another persistent connection to the database. By the start of 2016, the database had over 15,000 direct connections" https://t.co/Z5r72gXNZ1
— Joab Jackson (@Joab_Jackson) January 8, 2020
Something you rarely hear from the #K8s advocates: "Kubernetes has a model for how your infrastructure and services should behave. If you stray outside that model, then you'll be fighting k8s the entire way and it will be painful." Good convo on @newsycombinator https://t.co/3eBtBg7qSX
— Joab Jackson (@Joab_Jackson) January 14, 2020
"All of this glosses over the biggest issue with Kubernetes: it's still ridiculously complex … Compared to something like scp and restarting services, I would personally not pay the Kubernetes tax unless I absolutely had to."
— Joab Jackson (@Joab_Jackson) January 14, 2020
To run performance-sensitive apps on #Kubernetes, use the #K8S CPU Pools functionality to provide guaranteed isolation — & performance tuning — for the distinctly different types of workloads:@Nokia’s Levente Kale, Gergely Csatari @CloudNativeFdn #KubeCon https://t.co/2dLm1wh8oP pic.twitter.com/VtMGAaUrFo
— Joab Jackson (@Joab_Jackson) January 15, 2020
The engineers' message was clear: When dealing with a complex #Kubernetes-based infrastructure, performance tuning must be done all across the stack, including host, cluster, container, networking, and even with the underlying applications. https://t.co/srB6PnDnxM #KubeCon #o11y
— Joab Jackson (@Joab_Jackson) January 7, 2020
Data centers will devour up to 651 terawatt-hours of electric next year--nearly what Canada's entire energy sector produces. Streaming video, w/ 60.6 per cent of all internet traffic, is the biggest draw. "This will become completely unsustainable by 2040" https://t.co/MoPnwLH2ZM
— Joab Jackson (@Joab_Jackson) January 3, 2020
“Can I just once again state my love for it and hope it gets merged soon? Maybe the code isn’t perfect, but I’ve skimmed it, and compared to the horrors that are OpenVPN and IPSec, it’s a work of art”-- Linus Torvalds on the #WireGuard #VPN https://t.co/p3W1NVwdoD @thenewstack
— Joab Jackson (@Joab_Jackson) January 31, 2020
IBM lost a lot of momentum in the #HPC market after it sold its x86 business to Lenovo, and hasn’t kept momentum with its Power architecture. As we go into the #Exascale era of #supercomputing, AMD & ARM are among the up-and-comers. @ThisWeekInHPC #podcast https://t.co/CA4b08p3he
— Joab Jackson (@Joab_Jackson) January 27, 2020
The query that generated 300,000 alerts within a few hours and other @PrometheusIO cardinality disasters — @Bboreham @csmarchbanks #KubeCon https://t.co/yR582iVC7P pic.twitter.com/NBlhqhlOtN
— Joab Jackson (@Joab_Jackson) January 28, 2020
The perils of not timing a #Kubernetes upgrade correctly.. @kubecalc https://t.co/QaWdtw0Ynf
— Joab Jackson (@Joab_Jackson) January 30, 2020
A CNI driver connects the container to the network. Via REST, Kubernetes sends a JSON-packaged configuration to the driver, which provides a network interface within the container. @CloudNativeFdn maintains the CNI spec; 3rd parties build the drivers. https://t.co/kHSiVurHNl pic.twitter.com/4owSgJcQ1C
— Joab Jackson (@Joab_Jackson) January 31, 2020
"Deception and influence operations has been around for some time, but over the past several years, we have seen [...] companies grow up that basically build their business model around deception”--Facebook's Nathaniel Gleicher. https://t.co/kGriedxLLT
— Joab Jackson (@Joab_Jackson) January 7, 2020
“The design employed on the Death Star must allow the organic trash to filter down to the bottom, where the parasitic worm-creature devours it. But what happens when heavier pieces of non-organic trash fall down there?” https://t.co/qd0AtgNU2T
— Joab Jackson (@Joab_Jackson) January 7, 2020
E-mail us!!! 😂😂😂 https://t.co/BA4CplMDDz pic.twitter.com/vrZME9msI5
— Joab Jackson (@Joab_Jackson) January 4, 2020
I didn't even know you needed forklift training, & so I operated one for 8 months before I ran over something that pulled loose the transmission fluid hose (🙄), instantly bringing the unit to a halt, & leaking out a pool of dark red viscous-y liquid 😁 https://t.co/sLZKn56JJc pic.twitter.com/d6mXf07Hzh
— Joab Jackson (@Joab_Jackson) January 4, 2020
"The culture of tech pushes people harder and harder, but we don’t think about the physical effects of that labor. If we were athletes, where physical health is vital to success, the thinking would be completely different." @hairyelefante, @ozm https://t.co/u7WCUm4mhy
— Joab Jackson (@Joab_Jackson) January 16, 2020
“This is a toxic arrangement. The tactic shields tech companies from accountability. It allows giants like Amazon and Tesla an opportunity to transmit their preferred message, free of risk, in the voice of a given publication.” @cjr https://t.co/t8POnW7e3J
— Joab Jackson (@Joab_Jackson) January 23, 2020
snakey boi, obvs... https://t.co/QJlTJaMQ9Q
— Joab Jackson (@Joab_Jackson) February 1, 2020
I'm sorry to hear of Lars' passing. He was always immensely helpful -- and patient -- in explaining difficult concepts such as hypervisors and unikernels... https://t.co/y6oQW2obYH
— Joab Jackson (@Joab_Jackson) February 1, 2020