frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Autonomous recovery for distributed training jobs

https://docs.tensorpool.dev/features/agent
12•tsvoboda•1w ago
Hi HN! We’re TensorPool. We help companies access and optimize large scale compute for training foundation models.

The Problem

It’s been almost a year since we’ve finished YC, and we’ve just crossed 100,000 multinode training GPU hours run on our platform.

On those training runs, we’ve seen countless 3am job crashes because of issues like an Xid error from a flaky GPU or an S3 timeout that corrupted a checkpoint save. By the time you wake up and notice, you've lost 8+ hours of compute. You scramble to diagnose the issue, manually restart from the last checkpoint, and hope it doesn't happen again. Rinse and repeat.

For training runs that take days to weeks, this constant babysitting is exhausting and expensive. The research iteration cycles lost can also make or break a model release (especially for short reservations).

What We Built

This agent monitors your training jobs and autonomously recovers them when things go wrong. It works with Kubernetes, Slurm, and TensorPool Jobs.

We originally built the TensorPool Agent as an internal tool to help us debug failures with our own customers. Over time, we realized its performance was so good that we could automate the entire triage process. We're now releasing a public beta for people to use.

Best case: The TensorPool Agent detects the failure, diagnoses the root cause, fixes it, and restarts your job from the last checkpoint – all while you sleep ;)

Worst case: If the TensorPool agent can't fix the issue automatically, it delivers a preliminary RCA and a list of actions it attempted, giving you a head start on debugging.

How It Works

1) Registration – You provide credentials to your job scheduler via our dashboard. Perms are granted on a whitelist basis; you explicitly control what actions the agent can take.

2) Monitoring – The agent continuously monitors your job for failure conditions.

3) Recovery – On failure, the agent analyzes logs and attempts to diagnose the issue. If successful, it restarts the job from the last checkpoint and resumes monitoring. If not, you get an alert with full context.

Target Failure Modes

The agent is specifically designed for runtime errors that occur deep into training, like:

- CUDA OOM: Memory leaks, gradient explosions

- Xid errors: GPU hardware faults (Xid 79, 63, 48, etc.)

- Distributed communication failures: NCCL timeouts, rank failures

- Storage I/O errors: Checkpoint corruption

- Network issues: S3 request timeouts on mounted object storage

Comments

tsvoboda•1w ago
Would love to hear how you're handling recovery for long-running training jobs today, as well as what failure modes are most common/annoying for you.
hnotshe•1w ago
We're still figuring out how to detect "silent" failures where the job doesn't crash but stops making progress — like NCCL hangs where ranks are waiting indefinitely, or gradient norm explosions that don't trigger OOM but tank loss. Right now we rely on explicit errors in logs, but curious how others approach detecting "the job is technically running but something is very wrong" (if at all)?
jpollock•1w ago
Measurement and alerting is usually done in business metrics, not the causes. That way you catch classes of problems.

Not sure about expected loss, that's a decay rate?

But stuck jobs are via tasks being processed and average latency.

BTDUex Safe? The Back End Withdrawal Anomalies

1•aoijfoqfw•40s ago•0 comments

Show HN: Compile-Time Vibe Coding

https://github.com/Michael-JB/vibecode
1•michaelchicory•3m ago•0 comments

Show HN: Ensemble – macOS App to Manage Claude Code Skills, MCPs, and Claude.md

https://github.com/O0000-code/Ensemble
1•IO0oI•6m ago•1 comments

PR to support XMPP channels in OpenClaw

https://github.com/openclaw/openclaw/pull/9741
1•mickael•7m ago•0 comments

Twenty: A Modern Alternative to Salesforce

https://github.com/twentyhq/twenty
1•tosh•8m ago•0 comments

Raspberry Pi: More memory-driven price rises

https://www.raspberrypi.com/news/more-memory-driven-price-rises/
1•calcifer•14m ago•0 comments

Level Up Your Gaming

https://d4.h5go.life/
1•LinkLens•18m ago•1 comments

Di.day is a movement to encourage people to ditch Big Tech

https://itsfoss.com/news/di-day-celebration/
2•MilnerRoute•19m ago•0 comments

Show HN: AI generated personal affirmations playing when your phone is locked

https://MyAffirmations.Guru
4•alaserm•20m ago•3 comments

Show HN: GTM MCP Server- Let AI Manage Your Google Tag Manager Containers

https://github.com/paolobietolini/gtm-mcp-server
1•paolobietolini•21m ago•0 comments

Launch of X (Twitter) API Pay-per-Use Pricing

https://devcommunity.x.com/t/announcing-the-launch-of-x-api-pay-per-use-pricing/256476
1•thinkingemote•21m ago•0 comments

Facebook seemingly randomly bans tons of users

https://old.reddit.com/r/facebookdisabledme/
1•dirteater_•22m ago•1 comments

Global Bird Count

https://www.birdcount.org/
1•downboots•23m ago•0 comments

What Is Ruliology?

https://writings.stephenwolfram.com/2026/01/what-is-ruliology/
2•soheilpro•25m ago•0 comments

Jon Stewart – One of My Favorite People – What Now? with Trevor Noah Podcast [video]

https://www.youtube.com/watch?v=44uC12g9ZVk
2•consumer451•27m ago•0 comments

P2P crypto exchange development company

1•sonniya•41m ago•0 comments

Vocal Guide – belt sing without killing yourself

https://jesperordrup.github.io/vocal-guide/
2•jesperordrup•46m ago•0 comments

Write for Your Readers Even If They Are Agents

https://commonsware.com/blog/2026/02/06/write-for-your-readers-even-if-they-are-agents.html
1•ingve•46m ago•0 comments

Knowledge-Creating LLMs

https://tecunningham.github.io/posts/2026-01-29-knowledge-creating-llms.html
1•salkahfi•47m ago•0 comments

Maple Mono: Smooth your coding flow

https://font.subf.dev/en/
1•signa11•53m ago•0 comments

Sid Meier's System for Real-Time Music Composition and Synthesis

https://patents.google.com/patent/US5496962A/en
1•GaryBluto•1h ago•1 comments

Show HN: Slop News – HN front page now, but it's all slop

https://dosaygo-studio.github.io/hn-front-page-2035/slop-news
6•keepamovin•1h ago•1 comments

Show HN: Empusa – Visual debugger to catch and resume AI agent retry loops

https://github.com/justin55afdfdsf5ds45f4ds5f45ds4/EmpusaAI
1•justinlord•1h ago•0 comments

Show HN: Bitcoin wallet on NXP SE050 secure element, Tor-only open source

https://github.com/0xdeadbeefnetwork/sigil-web
2•sickthecat•1h ago•1 comments

White House Explores Opening Antitrust Probe on Homebuilders

https://www.bloomberg.com/news/articles/2026-02-06/white-house-explores-opening-antitrust-probe-i...
1•petethomas•1h ago•0 comments

Show HN: MindDraft – AI task app with smart actions and auto expense tracking

https://minddraft.ai
2•imthepk•1h ago•0 comments

How do you estimate AI app development costs accurately?

1•insights123•1h ago•0 comments

Going Through Snowden Documents, Part 5

https://libroot.org/posts/going-through-snowden-documents-part-5/
1•goto1•1h ago•0 comments

Show HN: MCP Server for TradeStation

https://github.com/theelderwand/tradestation-mcp
1•theelderwand•1h ago•0 comments

Canada unveils auto industry plan in latest pivot away from US

https://www.bbc.com/news/articles/cvgd2j80klmo
3•breve•1h ago•1 comments