@joabj.bsky.social
📤 90
📥 366
📝 5
Snowflake is buying Crunchy Data, an enterprise Postgres provider — Sridhar Ramaswamy,
#snowflakesummit
5 months ago
0
0
0
reposted by
The New Stack
7 months ago
An optimized version of Ray called RayTurbo will be available on Google Kubernetes Engine (GKE). Read more about this announcement from this article by
@joabj.bsky.social
loading . . .
Apache Ray Finds a Home on the Google Kubernetes Engine
Google Cloud and Anyscale have set out to make the Pythonic Apache Ray the de facto compute engine for AI and ML.
https://thenewstack.io/apache-ray-finds-a-home-on-the-google-kubernetes-engine/
0
2
1
reposted by
The New Stack
8 months ago
How Slack Transformed Cron into a Distributed Job Scheduler by
@joabj.bsky.social
loading . . .
How Slack Transformed Cron into a Distributed Job Scheduler
With help from Kubernetes, Golang and Kafka, Slack's crontab drives 2,000 tasks an hour. Monster Scale Summit had all the details.
https://thenewstack.io/how-slack-transformed-cron-into-a-distributed-job-scheduler
1
1
1
reposted by
The New Stack
7 months ago
Slack took cron from a single-server bottleneck to a distributed powerhouse handling 2,000 jobs per hour. With
#Kubernetes
,
#Golang
, and
#Kafka
, they built a resilient scheduler.
@joabj.bsky.social
unpacks how it all works.
loading . . .
How Slack Transformed Cron into a Distributed Job Scheduler
With help from Kubernetes, Golang and Kafka, Slack's crontab drives 2,000 tasks an hour. Monster Scale Summit had all the details.
https://thenewstack.io/how-slack-transformed-cron-into-a-distributed-job-scheduler/
0
2
1
I'm no MAGA hat, but this is a GREAT idea. F*ck pennies (and change in general) The NYC bodegas have the right idea. Something is either $0.50 or $1. Nothing in between is worth worrying about.
www.msn.com/en-us/politi...
loading . . .
MSN
https://www.msn.com/en-us/politics/government/trump-says-he-has-directed-treasury-to-stop-minting-new-pennies-citing-cost/ar-AA1yICXt?ocid=BingNewsSerp
9 months ago
0
0
0
Should federal CIOs be political appointees, where they can usher in change for each new administration, or should be career-long bureaucrats, defending the agency's long-term status quo?
fedscoop.com/opm-reclassi...
loading . . .
OPM pushes to reclassify chief information officers, opening up position to politics
The agency wants CIOs to be classified as “general” employees rather than “career reserved.”
https://fedscoop.com/opm-reclassify-chief-information-officers-career-politics/
9 months ago
0
1
0
reposted by
Rumman Chowdhury
9 months ago
I’ve spent about 2-3 hours each day on calls or chats with my US gov friends. Here’s an anonymized recap of what’s going on. First, I am so very sorry that they are living our last year at Twitter. I never wanted to think about that time again, and I hope my advice is helpful. 1/
51
1842
573
reposted by
The New Stack
9 months ago
@joabj.bsky.social
introduces CubeFS, the cloud-native file system built for Kubernetes. Optimized for storage/compute separation, it's a game-changer for databases, search systems, and AI/ML workloads in distributed environments.
loading . . .
Cloud Native Computing Now Has Its Own File System: CubeFS
CubeFS is especially attuned to the separation of storage/compute architecture for databases, search systems, and AI/ML applications.
https://thenewstack.io/cloud-native-computing-now-has-its-own-file-system-cubefs/
0
1
1
reposted by
Taylor Lorenz
9 months ago
My landlord is taking the opportunity to jack my rent up by hundreds of dollars a month (the maximum amount he’s allowed to) bc the LA wildfires have made rents skyrocket across LA. He knows I can’t afford to move so just have to accept it. Great system
65
1172
64
reposted by
The New Stack
9 months ago
Polar Signals couldn't figure out why it had so much expensive cross-zone traffic. So it turned to
#eBPF
to deliver the answers, and cut its cloud bill in half. By
@joabj.bsky.social
loading . . .
eBPF Tool Identifies Cross-Zone Kubernetes Network Traffic
Polar Signals couldn't figure out why it had so much expensive cross-zone traffic. So it turned to eBPF to deliver the answers, and cut its cloud bill in half.
https://thenewstack.io/ebpf-tool-identifies-cross-zone-kubernetes-network-traffic/
0
4
1
reposted by
Joey Hinson
9 months ago
Great article on what makes TikTok so special by
@joabj.bsky.social
He breaks down the recommendation engine and what makes it so special. Must read for anyone deep in the media world.
thenewstack.us11.list-manage.com/track/click?...
loading . . .
What Makes TikTok's Algorithms So Effective?
TikTok's recommendation system is incredibly good at understanding what users want — so good that it's the envy of tech titans like Elon Musk. But what makes TikTok tick?
https://thenewstack.us11.list-manage.com/track/click?u=ab6c02b160780b8e6569144f8&id=1c3f05c6e2&e=d7261a56a2
0
1
1
reposted by
The New Stack
10 months ago
After a 314-minute outage in 2022,
#Reddit
engineers reimagined their operations with platform abstraction.
@joabj.bsky.social
reveals how they tamed Kubernetes configurations.
loading . . .
Reddit No Longer Haunted by Drifting Kubernetes Configurations
Three years ago, Reddit's infrastructure engineers team spent most of its time putting out fires. Here's how developing a platform abstraction streamlined operations and put them back in charge.
https://thenewstack.io/reddit-no-longer-haunted-by-drifting-kubernetes-configurations/
0
2
1
reposted by
The New Stack
10 months ago
#IaC
tools scaled IT systems but left users grappling with Day 2 challenges.
@joabj.bsky.social
reviews 2024’s most-read IaC insights, highlighting potential relief in Infrastructure from Code. Do you think it's time to rethink DevOps infrastructure?
loading . . .
Infrastructure as Code in 2024: Why It's Still So Terrible
Tool fragmentation, integration hassles, configuration nightmares: Users of IaC tools are fed up. Could help come from the emerging practice of Infrastructure from Code?
https://thenewstack.io/infrastructure-as-code-in-2024-why-its-still-so-terrible/
0
2
1
reposted by
The New Stack
10 months ago
#Reddit’s
infrastructure overhaul turned chaos into control.
@joabj.bsky.social
details how the team used platform abstraction to conquer drifting Kubernetes configurations and streamline operations.
loading . . .
Reddit No Longer Haunted by Drifting Kubernetes Configurations
Three years ago, Reddit's infrastructure engineers team spent most of its time putting out fires. Here's how developing a platform abstraction streamlined operations and put them back in charge.
https://thenewstack.io/reddit-no-longer-haunted-by-drifting-kubernetes-configurations/
0
3
2
reposted by
The New Stack
12 months ago
The "MultiKueue" beta multicluster job dispatching feature allows admins to place workloads on remote clusters. By
@joabj.bsky.social
loading . . .
Kueue Can Now Schedule Kubernetes Batch Jobs Across Clusters
The "MultiKueue" beta multicluster job dispatching feature allows admins to place workloads on remote clusters.
https://thenewstack.io/kueue-can-now-schedule-kubernetes-batch-jobs-across-clusters/
0
2
1
Unlike traditional high-performance computing (HPC) systems, CoreWeave does not run on services on bare metal but rather uses Kubernetes over the bare metal.
thenewstack.io/hpc-kubernet...
#Kubecon
#CloudNativeCon
loading . . .
HPC Kubernetes: AI Training on 3,500 GPUs
K8s brings many advantages to managing fleets of GPUs, said CoreWeave's Peter Salanki, during a talk at KubeCon+CloudNativeCon 2023.
https://thenewstack.io/hpc-kubernetes-ai-training-on-3500-gpus/
12 months ago
0
1
0
“We have to band together and figure out how to deal with this problem”— Chris Aniszczyk on the increasing problem of Patent Trolls for
#Opensource
#CloudNative
technologies.
#Kubecon
#CloudNativeCon
12 months ago
0
1
0
you reached the end!!
feeds!
log in