Metadata

This Berkeley systems group paper opens with the thesis that LLM agents will soon dominate data system workloads. These agents, acting on behalf of users, do not query like human analysts or even like the applications written by them. Instead, the LLM agents bombard databases with a storm of exploratory requests: schema inspections, partial aggregates, speculative joins, rollback-heavy what-if updates. The authors calls this behavior agentic speculation . Agentic speculation is positioned as both the problem and the opportunity. The problem is that traditional DBMSs are built for exact intermittent workloads and cannot handle the high-throughput redundant and inefficient querying of LLM agents. The opportunity also lies here. Agentic speculation has recognizable properties and features that invite new designs. Databases should adapt by offering approximate answers, sharing computation across repeated subplans, caching grounding information in an agentic memory store, and even steering...

This short VLDB'25 paper  surveys disaggregation for cloud databases. It has several insightful points, and I found it worth summarizing.  The key advantage of the cloud over on-prem is elastic scalability: users can scale resources up and down and pay only for what they use. Traditional database architectures, like shared-nothing, do not fully exploit this. Thus, cloud-native databases increasingly adopt disaggregated designs. Disaggregation is primarily motivated by the asymmetry between compute and storage: Compute is far more expensive than storage in the cloud. Compute demand fluctuates quickly; storage grows slowly. Compute can be stateless and easier to scale, while storage is inherently stateful. Decoupling them lets compute scale elastically while storage remains relatively stable and cheap. Review of Disaggregation in the Clouds Early cloud-native systems like Snowflake and Amazon Aurora separate compute and storage into independent clusters. Modern systems push dis...

This year was a bastard. Not from work. God, no, I find work relaxing. Reading papers, inventing algorithms, ripping apart distributed systems with TLA+ models -- that's how I have fun. I can do that all day with a grin on my face. But the minute I need to do adulting (like simply calling the cable company and ask why keep increasing our bill when I'm not looking), I will stress and procrastinate. And this year, I had a lot of adulting to do to put our house on market, and plan a move to California, all the while juggling to help three kids with school and EC activities. No wonder I've been grinding my teeth at night like a mule chewing rocks... Anywho, we botched our great escape to California. House didn't sell quickly, as we hoped it would, and we are stuck in Buffalo for another year. Summer disappeared in a cloud of errands and disappointment, and suddenly it was late August with our kids twitching with pre-school nerves. There was still some time left to salvage t...

Small Gods (1992) I absolutely loved Small Gods. Pratchett takes on religion, faith, and power. The story follows Om, a god trapped in the body of a tortoise, who has only one true believer left: a novice named Brutha. The central premise is that gods and mythical beings exist because people believe in them, and their power fades as belief fades.      There’s no point in believing in things that exist. The book is funny, clever, and surprisingly philosophical. Pratchett skewers organized religion, but he also asks bigger questions: What is faith? What is belief? How do institutions shape people, and how do people shape institutions? It's satire, but not heavy-handed. Like Vonnegut, he writes with a wink, yet there's real depth under the jokes.      Gods don't like people not doing much work. People who aren't busy all the time might start to think.      Why not? If enough people believe, you can be god of anything….      The f...

People want data fast. They also want it consistent. Those two wants pull in opposite directions. This VLDB'25 paper does another take on this conundrum. Rather than assuming a symmetric network environment where all replicas face similar latencies, the paper emphasizes that in practice, some replicas are closer to the leader, where others are stranded halfway across the globe. By embracing this asymmetry, the authors propose two new algorithms: Pairwise-Leader (PL) and Pairwise-All (PA). Both cut read latency compared to the prior approaches. PL could even achieve 50x latency improvements in some cases. Aleksey and I did our usual thing. We recorded our first blind read of the paper . I also annotated a copy while reading which you can access here. We liked the ideas, even though the protocols themselves didn't thrill us particularly. I particularly liked finding another good example of the use of synchronized time in distributed database systems. Another study to add to my s...