ceph
Ceph RGW Accounts in Tentacle: Plan the Identity Model
Ceph tenants and accounts solve different problems. Separate the Tentacle upgrade from account adoption, then test ownership, IAM and notifications.
data
Data Platform Resilience: RTO, RPO and Replication in Practice
A DR document states minutes and megabytes. Replication architecture delivers a convergence time, a single-copy window and a failed tail. This guide maps what actually decides your RTO and RPO.
data
RTO and RPO for Data Platforms: Test the Recovery Path
RTO and RPO are measurable recovery objectives, not labels on an architecture diagram. Build a failure matrix and test the complete recovery path.
object storage
Object Storage for Data Platforms at Scale
A production guide to running object storage at scale: why object count is a capacity axis, how prefix design and the background scanner set your real SLAs, and how to choose engines, media and release channels.
data
Benchmark Best Practices, Part 2: Little’s Law Ties Throughput, Latency and Concurrency Together
Second in my benchmark best practices series. Little’s law checks that throughput, latency and concurrency describe the same system, exposes the load generator’s ceiling, and turns a latency budget into a sizing requirement.
MinIO
MinIO delete-marker convergence: how the Silo fork fixed the 405 replication loop
The 405 storm had three root defects in the delete-marker path. PR #162 fixed the first attempt; PR #184 fixes retries, purge status and MRF healing in the Silo fork.
MinIO
MinIO Community in Production: Assess the Risk You Are Running
MinIO Community Edition has had no release since October 2025 and the repository is archived. Three CVEs will never be patched there. Here is how to assess what your cluster is actually exposed to, and where the maintained community fork fits in.
AI
Your AI Supply Chain Is a Production Dependency. Audit It Like One.
Two movements meet: AI is getting production responsibilities, and production data is flowing through AI. Anthropic’s September report, malicious routers, training data provenance and model drift show what to verify before an AI reaches production.
MinIO
MinIO site replication bug: how 684 million HTTP 405s exposed the delete-marker loop
HEAD probes on delete markers answer 405 and replication counts that as success. 20M stuck versions turned the loop into 684M useless requests in 48 h.
AI
The First Vibe Coding Data Is In, and It Complicates DHH’s Claim
A preregistered ETH Zurich study found that computer science knowledge, not writing skill, is the strongest predictor of vibe coding success. Here is what that does and does not say about DHH’s claim that programmers can be worse at AI coding than people who cannot code.