Keeping CALM: When Distributed Consistency is Easy

Interesting that neither the article nor the comments mention the CALM theorem [0], which gives a framework to explain when coordination-free consistency is possible, and is arguably the big idea behind SEC. [0] https://arxiv.org/abs/1901.01930

charleshnSep 10, 2025View on HNParent context

> Distributed algorithms require more coordination. Sometimes! There's a whole body of research about when distributed algorithms require coordination. One example is the CALM theorem (https://arxiv.org/abs/1901.01930), and another is the ways that scalable database systems…

mjbFeb 13, 2025View on HNParent context

Yes it's a large chunk, but not everything! Marc had a comment on bluesky regarding this: > Many SQL aggregations are monotonic operations (e.g. MAX, SUM, etc) that can be partially completed on each node and then post-merged. Some (e.g. DISTINCT) can be transformed into…

steepbenDec 4, 2024View on HNParent context

> Perhaps others have made this connection before…it does feel like some sort of universal law of physics. This has been on my mind too, and I can’t help but think the fundamental concept underpinning it all is causality. Reading about “consistency as logical monotonicity” —…

ojkellyOct 8, 2024View on HNParent context

don't forget about calm. its a shame that that isn't the default we reach for, and only struggle when we really need stronger latency. https://arxiv.org/abs/1901.01930

convolvatronSep 27, 2024View on HNParent context

*Between two distributed systems that don't share the same transactionality domain or are not logically monotone. It's easy to see that moving data from one row to another is doable in a clustered database, and could be interpreted as a message being delivered. The point is that…

j-pbSep 3, 2024View on HNParent context

The article should really mention CALM (Consistency as Logical Monotonicity)[1], it's much easier to understand and a more fundamental result than CAP. It is also much more applicable and enables people with little experience to build extremely robust distributed systems.…

j-pbSep 3, 2024View on HNParent context

There is a theorem called CALM[1] (Consistency as Logical Monotonicity) that shows that systems are eventually consistent iff they are monotonic. I find that reasoning about consistency is much easier from that perspective, for example it immediately gives you an intuition on…

j-pbMar 8, 2024View on HNParent context

Completely agree. In-region (e.g. at the scale of a US state), CP seems like a clear winner. For more geo-distributed latency becomes challenging, and in environments like IoT and mobile unreliable connectivity becomes challenging. If you're there, you need a principled way of…

mjbJan 31, 2024View on HNParent context

It's absolutely bonkers that TLA+ struggles with this, given that a system is eventually consistent IFF it is logically monotonic (CALM) [1]. This also doesn't make intuitive sense to me. Model checking should be easier for monotonic models, because because the total model check…

j-pbNov 2, 2023View on HNParent context

> How do you build a system around a database where values can no longer be valid at any time? This is a difficult problem, and there aren't nice general solutions. If you read the original Amazon Dynamo paper, for example, there's a good discussion of why eventual consistency…

mjbJun 7, 2023View on HNParent context

yeah, that was really brief. I was actually hoping for someone to bring up https://arxiv.org/abs/1901.01930 (keeping CALM) for an interesting alternative

convolvatronApr 6, 2023View on HNParent context

You might be thinking of this [1] paper? The core idea being if your problem space lends itself to monotonicity (see the paper for a more precise definition than I can give), then you can build a globally consistent database (around a CRDT) where the end-user doesn't need to…

newhousebNov 27, 2022View on HNParent context

Are you familiar with Joe Hellerstein and Peter Alvaro's work in this space? CALM provides something of a unifying theory of mergeable operations, of which CRDTs are an "object orientated" special case: https://arxiv.org/pdf/1901.01930.pdf In practice, designers choose…

mjbFeb 11, 2021View on HNParent context

> That function can do whatever it wants, including any number of reads, writes, and arbitrary logic based on the reads. But critically, it can't talk back to the calling client, except for to send the final result. If that function can still do everything the client did and the…

zzzcpanNov 8, 2019View on HNParent context

There is something based on Bloom IIRC, but the whole concept should be well-described and discoverable from the Keeping CALM paper: https://arxiv.org/abs/1901.01930

namibjMay 28, 2019View on HNParent context