#distributedsystems — Public Fediverse posts

Michael Pigott on Toward a Generic Fault Tolerance Technique [PWL NYC] (2021)

https://www.youtube.com/watch?v=W55mk2CivJo

#computerscience #distributedsystems #networks

Shorter loops, stronger proof.

Reports → reviewed changes. Commits → independent mirrors. Faster, clearer, harder to mistake for finished work before the proof is there.

AI-generated; occasional AI posts.

#ForkMesh #OpenSource #FOSS #DevTools #Git #SelfHosted #Decentralized #DevOps #Programming #Developers #Coding #GitOps #Infra #DistributedSystems #PeerToPeer #Tech #Engineering #OSS #CodeReview #IndieWeb #Federation #SoftwareFreedom #DigitalSovereignty #BuildInPublic
https://forkmesh.com/blog/shorter-loops-stronger-proof/

This week at SIGCOMM, Inho Choi will present a paper about a fast TCP migration and load balancing system called Capybara, which means it’s time for… another episode of Distributed Systems Explained by ̶C̶a̶t̶s̶ Capybaras!

L4 load balancers are fast, scalable, and widely used in cloud systems: when a new connection arrives, the load balancer picks a server for it. For our purposes, connections are birds and servers are capybaras.

This bird has been assigned to a server.

[🧵, 1/9]

#distributedsystems #networking #capybara #sigcomm #distributedsystemsexplainedbycats

PWLSF - 08/2014 - Peter Alvaro: Using Reasoning about Knowledge to Analyze Distrtibuted Systems (2014)

https://www.youtube.com/watch?v=7U0qPmEpbSI

#ReasoningAboutKnowledge #distributedsystems #Software

This pattern also has to plan for cancellation, retries, and cleanup, not just the happy path of request, poll, and result.

#Fintech #DistributedSystems

On local networks, Fire-and-Forget messages can travel over UDP, while most messaging systems, such as JMS-compliant brokers, provide this style by default.

#Fintech #DistributedSystems

How Netflix redesigned its Service Topology pipeline to scale real-time service mapping:
🔹 Separates resolution from enrichment & persistence
🔹 Propagates backpressure to Kafka (zero record drops)
🔹 Swaps gRPC for SSE on high-volume internal transfers

📖 Read the full deep dive on #InfoQ ⇨ https://bit.ly/4x0Hrzc

#DistributedSystems #SoftwareArchitecture #Kafka #Microservices

ATProto for Distributed Systems Engineers

https://atproto.com/articles/atproto-for-distsys-engineers

Comments: https://news.ycombinator.com/item?id=49202912

#HackerNews #ATProto #DistributedSystems #Engineers #TechInnovation #SoftwareDevelopment #Networking

What I’ve been reading (and watching) week ending 9 August https://jchyip.medium.com/what-ive-been-reading-and-watching-week-ending-9-august-1ab6b2dde859 #AI #DistributedSystems #science #engineering #SoftwareEngineering #cybersecurity #ProductManagement

John Valois on Wait-Free Synchronization [PWL NYC] (2019)

https://www.youtube.com/watch?v=7end3rQ0jkk

#distributedsystems #tech #software

PWLSF- 09/2014 - Armon Dadgar on SWIM (2014)

https://www.youtube.com/watch?v=aVYEY75qn3c

#distributedsystems

Apache Kafka concepts can be tricky; visualizing them helps.

The SoftwareMill Kafka Visualization tool breaks down the fundamentals interactively so you can see how things work in real time.

🆕 Now updated to support Kafka Share Groups.

Give it a spin: https://softwaremill.com/kafka-visualisation/

#ApacheKafka #Kafka #DistributedSystems #SoftwareEngineering

Cloudflare has introduced Meerkat - a globally consistent control-plane service built on the QuePaxa consensus algorithm.

By enabling leaderless writes while maintaining strong consistency, it improves availability & resilience across Cloudflare's global network.

Read the full story on InfoQ 👉 https://bit.ly/4q0xxLg

#DistributedSystems #Cloudflare #Consistency #Availability #InfoQ

Asynchronous designs trade immediate answers for resilience: a call can be retried until it succeeds, but the sender no longer knows the outcome right away.

#Fintech #DistributedSystems

Synchronous communication means a caller sends a request and blocks until the receiver replies. It exists so a service can get an immediate, ordered answer before it moves to the next step.

#Fintech #DistributedSystems

I've had a delightfully nerdy morning - I've made it so that #fishnet ("distributed Stockfish analysis for lichess.org", https://github.com/lichess-org/fishnet), runs on my computer ONLY when the temperature is below 25°C and the grid is powered by more than 40% renewable energy. Very silly project, but FUN 🤓 🌞

It uses https://data.hub.api.metoffice.gov.uk/sitespecific/v0/point/hourly?excludeParameterMetadata=true&latitude=${LAT}&longitude=${LON} for the temperature, and https://api.carbonintensity.org.uk/generation for the % of renewable energy on the grid.

I think the thing I'm most looking forward to is that it will be an abstract way to keep an eye on how the grid is doing (e.g. "Ooh, my computer fans just kicked in, the grid must be pretty green today! 🎉 💚 ")

#lichess #chess #DistributedComputing #DistributedSystems #GreenComputing

From the Leanpub Blog: Leanpub Book LAUNCH 🚀 Architecture Synthesis: The Next Correct Step by Sergiy Yevtushenko

https://leanpub.com/blog/leanpub-book-launch-architecture-synthesis-the-next-correct-step-by-sergiy-yevtushenko/

#books #leanpublishing #selfpublishing #softwarearchitecture #systemdesign #architecturedecisions #backendengineering #distributedsystems

NEW! Leanpub Book LAUNCH 🚀 Architecture Synthesis: The Next Correct Step by Sergiy Yevtushenko

https://youtu.be/mkH6U6Xws6g

#books #leanpublishing #selfpublishing #softwarearchitecture #systemdesign #architecturedecisions #backendengineering #distributedsystems

DBOS CEO Jeremy Edberg & co-founder Qian Li explain how DBOS Transact runs durable workflows straight inside #Postgres - using just two tables, ✅ a SKIP LOCKED queue, and ✅ zero external orchestrators.

🎬 Watch the #InfoQ video & read the full transcript here: https://bit.ly/45qo6ei

#Database #SoftwareArchitecture #DistributedSystems #OpenSource

Learn everything you need to know about Distributed Systems via these 257 free HackerNoon blog posts. https://hackernoon.com/257-blog-posts-to-learn-about-distributed-systems #distributedsystems

RE: https://mastodon.social/@koen_hufkens/116979524452174787

This is absolute insanity.
We could have decentralized index and no reason to crawl websites like madman.
We have search engines that crawl the web regularly, including Google, and they can behave!

But ofc people in silicon Valley building shit mentality prefer the hammer style. Brute forcing everything.

See OpenAI and anthropic bellow. That is just evil, it should be considered a DDOS and punishable by law.

What is the solution?
We can't avoid what they are doing right now, the PandoraBox is open, but what can we do?

I wonder if we could get traffic shapes to tell us if this is coming from harnesses or directly from different LLM vendors such as OpenAI, Anthropic, ZAI, etc.
This website seems to be facing the later, but traffic from harnesses are still egregious already.

If the traffic come from vendors we should sue them for DDOS, if it's from harnesses, I think many of them are open source we could contribute so they start using a distributed index or existing search engine results.

Any other ideas?

#search #distributedsystems #openai #anthropic
#searchengine #webcrawler #webcrawling

DoorDash's new Entity Cache runs on Envoy and Valkey, caching over 100 endpoints across 50 services at 1.5M requests/sec with 99.99999% availability.

Using Kafka-driven invalidation and the XFetch algorithm, it reduces cache stampedes and cuts P99 latency spikes by up to 80%.

Read the full story on InfoQ 👉 https://bit.ly/45jwGvi

#Caching #Envoy #Valkey #Kafka #DistributedSystems #InfoQ

ClouDens: Operational Context-Aware Anomaly Detection for Large-scale Cloud System Monitoring

https://arxiv.org/abs/2607.18127

#Monitoring #DistributedSystems