mjb

31 ms·

271.
by mjb 6y ago
Sharding is a great way to scale, but comes with costs. One cost is that you need a separate transaction layer "on top of" the shards if you want to offer transactions. If you want to offer strong isolation (such as snapshot or se
272.
by mjb 6y ago
As others have said, TLA+ can be very useful outside distributed systems. For example, I'm currently working on a virtualization-related low-level systems paper (likely to be submitted to VEE) where we use TLA+ to reason about the safe
273.
Raft does not guarantee liveness in the face of network faults(decentralizedthoughts.github.io)
2 points by mjb 6y ago|0 comments
274.
by mjb 6y ago
Looks like a great talk. I'm looking forward to watching the whole thing later. Jumping forward to the TLA+ section, I very much like the point about how TLA+ enabled precise communication between teams. Writing about distributed algor
275.
by mjb 6y ago
This is excellent stuff, as is all of Peter's (and others) prior work in this area. It's a great step from "you can't do X" (like https://groups.csail.mit.edu/tds/papers/Lynch/MIT-LCS-
276.
by mjb 6y ago
> Which is directly at odds with needing TLA+, because if you do need it, it means the complexity of the algorithms is so great, that you won't be able to keep them in your head and understand every aspect of their performance to ma
277.
by mjb 6y ago
You don't need to build from a Lambda specific base image - you can use (nearly) any base image you choose. Your application just needs to implement the Lambda runtime API, which we provide clients for in multiple languages.
278.
by mjb 6y ago
Fun article! I really don't like that diagram. The folks doing serverless work at University of Wisconsin-Madison generally do a good job, but they really whiffed on that one. I don't think there's a reasonable sense in whi
279.
by mjb 6y ago
This taught me about ahash ( https://crates.io/crates/ahash ). It uses the aesenc ( https://www.felixcloutier.com/x86/aesenc ) instruction as a high-quality fast permutation. Smart and fun.
280.
by mjb 6y ago
This is a really good post. I liked seeing the TLA+ code here. There's a pretty serious ambiguity in the Paxos Made Simple paper, which makes it very easy to follow the paper and still end up with a broken algorithm. From Lamports publ
281.
by mjb 6y ago
> While it increases the data locality, I have seen a few software following this sharding model (notably scylla) that work really bad once the load is not evenly distributed across all shards Serial access to hot keys is a hard thing to
282.
by mjb 6y ago
This article, dated 6th November 2013 was clearly posted here to have a particular discussion, and to push a particular political agenda. It was posted by somebody who posts only right, conservative, and white-supremacist talking points to
283.
by mjb 6y ago
Hey, this is my blog post! Great to see it here. Best-of-two is one of my favorite things. If you enjoy the simplicity of it, check out the related paper "How asymmetry helps load balancing" by Berthold Vöcking ( https:/&#x2F
284.
by mjb 6y ago
> I did my undergrad CS thesis on this topic and implemented a "power of two choices" load balancing module in NGINX. Cool! > When simulating load distribution across servers, two-choices does an outstanding job at keeping t
285.
Getting Big Things Done(brooker.co.za)
2 points by mjb 6y ago|0 comments
286.
by mjb 6y ago
I'm also biased here, but it seems like Firecracker ( https://github.com/firecracker-microvm/firecracker ) and rust-vmm ( https://github.com/rust-vmm ) would be great things to use as examples in a co
287.
by mjb 6y ago
Teaching a class about safety in systems code seems like a great idea, and at first glance the class content here seems useful and interesting. I do a lot of OS-level stuff at work (especially around virtualization and Firecracker), and whi
288.
by mjb 6y ago
But sparks and arcs aren't impulses. They're much more interesting than that spectrally, including with significant peaks and harmonic components. These properties are things that protection equipment can use to detect and localiz
289.
Consensus is harder than it looks(brooker.co.za)
3 points by mjb 6y ago|0 comments
290.
by mjb 6y ago
Nice to see ZGC marked as production-ready. With Shenandoah, ZGC, and G1 the options for garbage collection in Java are so much better than they were just a couple of years ago. Even better, the need to do a bunch of brittle tuning and twea
291.
by mjb 6y ago
> One of the interesting things about Raft that doesn't get focused on much is that strictly speaking Raft does not require that all nodes be able to see each other, just that some nodes can see all of the other nodes. The same is t
292.
by mjb 6y ago
Very cool post. I like this point particularly: > In this case we can’t assume, in general, that the thread scheduler will assign each thread processor time with uniform probability. That assumption makes even less sense in a fully distr
293.
by mjb 6y ago
> Newer CPUs, e.g.: AMD EPYC or any recent Intel Xeon can do about 1 GB/s/core, but I haven't seen any CPUs that significantly exceed that. Modern processors can do ~2.1GB/s per core for AES-256-GCM (about 2x what the
294.
The State of the Serverless Art (2020)(medium.com)
1 points by mjb 6y ago|0 comments
295.
by mjb 6y ago
Right, 'asymptotic' is a better description of the behavior. I've updated the post.
296.
by mjb 6y ago
Downloading dependencies on cold start not only adds to start latency, but is also an availability risk. You don't want your starts to fail when the package distribution infrastructure is down. Where possible, it's best to include
297.
by mjb 6y ago
Thanks for sharing that, I'll read it! I do agree that comments, documentation, and other artifacts aren't sufficient to solve this problem. The closest I've come is with formal specification, where the intent of the program
298.
Code only says what it does(brooker.co.za)
170 points by mjb 6y ago|114 comments
299.
Code only says what it does(brooker.co.za)
1 points by mjb 6y ago|0 comments
300.
by mjb 6y ago
One funny thing about that VLDB'12 paper is Kevin Grittner's affiliation. Not often you see a state supreme court publishing cutting-edge database research. I'd love to hear the story around that.

More ›