Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Oxide Computer: Docs

docs.oxide.computer

Post reply on HN
Print view View original

141–150 of 160 posts

Re: Oxide Computer: Docs

#141

Post by gorkish »

javajosh wrote: post #73
soneil wrote: post #49

Earlier quoted context omitted.

That's pretty standard? Anything that's going straight into 400V (I assume) deserves a good eyeball before it's energised. I doubt their "root of trust" knows how to protect the system from a dead mouse in the bus bars.

I'm not worried about dead mice; I'm worried about tampering. I would have hoped there'd be more guidance than "just look and see if anything looks weird". A determined threat actor can easily side-step that check with, for example, careful desoldering/resoldering technique.

If this is seriously your level of supply chain scrutiny, being an Oxide early adopter probably isn't the best idea.

Re: Oxide Computer: Docs

#142

Post by gorkish »

As someone who has been attempting to smash computers together into 'hyperconverged infrastructure' since before Y2K, I could not be happier for the buzz around Oxide; hopefully we'll see smaller scale versions of the concept out soon -- something in the 3-10 node size for SMB and/or a small scale-out system for SOHO. It seems insane to me that the tech to do this with off the shelf free software exists, but there's no way to buy it ready to go at small scale. Go get a quote for an HCI vmware buildout and DISMAY.

Re: Oxide Computer: Docs

#143

Post by Tuna-Fish »

soneil wrote: post #49
javajosh wrote: post #19

I'm particularly impressed with their anti-tamper measures. "For each server sled, shine a light into the cubby to look for any physical tampering or damage." https://docs.oxide.computer/guides/system/rack-installation-...

That's pretty standard? Anything that's going straight into 400V (I assume) deserves a good eyeball before it's energised. I doubt their "root of trust" knows how to protect the system from a dead mouse in the bus bars.

> Anything that's going straight into 400V (I assume)

I think the sleds blind mate to a 55V DC common bus.

Re: Oxide Computer: Docs

#144

Post by realitythreek »

gorkish wrote: post #134
LoganDark wrote: post #26

Earlier quoted context omitted.

Are these available in text or article format?

O+F is posted to youtube, I get the auto-generated transcript from there. It's unscripted/unstructured, so if you are primarily wanting to hear discussion of the oxide hardware/software/progress it's pretty difficult to consume as an audio program. This is not a dig on the program at all; I'm glad they are making the time to produce it, and I'd rather they spend their effort getting racks out the door instead of gene…

I always miss the live Discord call and never realized they were published to YT! Thanks for the notice.

Re: Oxide Computer: Docs

#145

Post by heudhdidh »

petesergeant wrote: post #116
bcantrill wrote: post #102

Earlier quoted context omitted.

This conversation strikes me as unlikely on several levels. First, no one would have coached you on "how to interview at Oxide" because that's not where the process starts -- it starts with you preparing your materials.[0] (Our review of the materials constitutes ~95% of our process.) Second, we have always been very explicit about compensation (that is, we ourselves brought it up early in conversations); no one at O…

Hey Bryan, I enjoyed reading the cash compensation article, but I'm curious about how it meshes with your equity compensation? Is it awarded purely on when people joined? Does everyone who wasn't a founder get the same amount? It feels like you could run into much of the same problems the article points out about transparency and so on if the equity component isn't as straight-forward?

speaking as someone who dropped on the first round: for the nobodies like me, they say equity is based on time joining. it was early 2021 i think and amount already published on the job or mentioned pretty early (don't really recall)... and it was brutally low considering the pay cut. maybe they keep it to negotiate the somebodys.

Re: Oxide Computer: Docs

#146

Post by actionfromafar »

mst wrote: post #136
actionfromafar wrote: post #123

Earlier quoted context omitted.

They could probably have someone proofread the transcripts, there aren't that many episodes.

I was going for "set up some code once and don't think about it again" to maximise the odds of the idea sounding tempting. Proofreading would set up an expectation on the part of readers that it -had- been proofread and corrected and therefore a commitment to perform a repeated "boring but important" task going forwards for whoever's doing said proofreading. That way would likely lie either delayed transcripts or nev…

I hate bad transcripts.

Re: Oxide Computer: Docs

#147

Post by yencabulator »

kaliszad wrote: post #119
drbawb wrote: post #38

I am super interested in learning more about the storage subsystem! I figured they'd be using ZFS, given the people involved, but it appears they've also gone ahead and built a clustered FS (crucible) on top of it? I figured something like that would be necessary to handle fault tolerance at the gimlet level. (Losing an entire shelf / drive controller, etc.) Getting ZFS to go multi-node is surely a neat trick. Second…

They use Crucible on top of ZFS. https://github.com/oxidecomputer/crucible I don't think they have anything for S3-like service but there are other options for that, e.g. https://garagehq.deuxfleurs.fr or MinIO. I am not sure whether they have their own SSDs or use of the shelf SSDs just with their firmware or something.

They implemented a block store with replication from scratch? That's kinda brave, considering that that's a project big enough to justify full startups for!

Re: Oxide Computer: Docs

#148

Post by yencabulator »

throwaway2037 wrote: post #94
wmf wrote: post #53

Earlier quoted context omitted.

I would say Oxide is inflexible and non-customizable since they have exactly one hardware configuration and few software features at this point. Their claim is more that their rack works and everything else on the market is full of bugs.

I'm not an infra engineer, but this claim "everything else on the market is full of bugs" might be the killer app. Of course, it needs to be true. What if they iterate to an insanely stable embedded code base (BIOS, etc.)? Then, continuously upgrade the hardware to use the latest CPU/RAM/NVME. I could see that being very valuable.

I fully expect there will be data-loss bugs and poor performance during recovery in their in-house distributed block storage solution. That's just in the nature of the problem domain, and this is all new code:

https://github.com/oxidecomputer/crucible

Re: Oxide Computer: Docs

#149

Post by yencabulator »

codetrotter wrote: post #6

I wonder how much the smallest, cheapest configuration will cost. I would really love to buy one of these. But I suspect it will cost 100x more than I can afford heh.

Supermicro equivalent to what one of those sleds might be is around $14k. Times 32, plus switches etc, plus proprietary development margin, plus lack of volume manufacturing benefits.

https://www.siliconmechanics.com/system/rackform-a235.v8.1/2...

Re: Oxide Computer: Docs

#150

Post by yencabulator »

javajosh wrote: post #73
soneil wrote: post #49

Earlier quoted context omitted.

That's pretty standard? Anything that's going straight into 400V (I assume) deserves a good eyeball before it's energised. I doubt their "root of trust" knows how to protect the system from a dead mouse in the bus bars.

I'm not worried about dead mice; I'm worried about tampering. I would have hoped there'd be more guidance than "just look and see if anything looks weird". A determined threat actor can easily side-step that check with, for example, careful desoldering/resoldering technique.

I think for the unit price of $1M per rack or more, you can easily afford an armed guard to escort the crate from the factory.
Post reply on HN