If you declare your business logic in the database, the SQL-92 standard specifies SQL assertions, but no SQL databases have implemented them before 2026. Here is an example using @martin.kleppmann.com's example (Doctor's On-Call Shift) on Oracle 23.26.1 dev.to/franckpachot...

My first test of CloudNativePG: dev.to/franckpachot...

Once won’t hurt — I did a silly benchmark with a clickbait headline 🤣: #PostgreSQL is faster when replacing SQL tables with MongoDB collections dev.to/franckpachot...

When enterprises evaluate database alternatives, Oracle MAA is the reference for high availability, as it runs many banks and hospitals. Here’s how MongoDB achieves similar HA objectives: dev.to/franckpachot...

The 50-year quest to replace developers ❌ 1970: COBOL ❌ 1980: CASE tools ❌ 1990: Visual Basic ❌ 2000: Low/No-Code ❓ 2020: #AI Result? We need more skilled developers than ever. Blocker isn't the code; it's the complexity. 🧑‍💻💡 By Stephan Schwab www.caimito.net/en/blog/2025...

🌱🐘 #PostgreSQL or MongoDB? - the recording of my talk in Madrid youtu.be/HG4HEK95SbU?...

💡MongoDB stores objects in BSON, but from the application, you access fields with a Document object created by the driver—don't use RawBsonDocument, which is just an array of bytes. All details in a new blog post: dev.to/franckpachot...

MongoDB equivalent to FOR UPDATE SKIP LOCKED SELECT ... FOR UPDATE SKIP LOCKED is a vendor-specific SQL feature available in several relational databases (e.g., PostgreSQL, Oracle, MySQL). It helps... #mongodb #postgres #database #lock Origin | Interest | Match

MongoDB's findOneAndUpdate() returns the document updated (before or after image). Bonus: MongoDB makes it idempotent, making it resilient to failures. A new blog post: dev.to/franckpachot...

If you’ve read my last post on read and write concerns, you might wonder why MongoDB’s default read concern isn’t "majority, so I've written a new blog post: dev.to/franckpachot...

Once you understand your access patterns, you can tune your MongoDB data model, indexes, and queries. dev.to/franckpachot...

Comparing #MongoDB Read/Write Concerns and #PostgreSQL Synchronous Commit Replication in terms of performance, durability, visibility, and isolation: dev.to/franckpachot...

Sequence diagrams are overlooked. Here's an example of a replication sequence with w:majority in a #MongoDB PSS deployment. The horizontal axis (actors) shows the performance–durability trade-off, and the vertical axis (timeline) shows the visibility–isolation trade-off:

First blog post of the year - a multi-thread pick-process-update. Avoid calling external APIs within one transaction. Manage the state in the document with short updates: dev.to/franckpachot...

MongoDB Read/Write vs. PostgreSQL Synchronous Replication dev.to/franckpachot...

There are no Foreign Keys in MongoDB because strong one-to-many relationships are embedded, and data integrity is handled in the application via business logic. You can also validate database integrity using the aggregation pipeline and the change stream: dev.to/franckpachot...

A new blog post about an @awscloud.bsky.social Amazon Aurora PostgreSQL feature that exists in many commercial databases but not in PostgreSQL: adaptive join methods dev.to/aws-heroes/a...

New blog post: examining an Amazon Aurora query planner transformation and manual methods, with notes on COUNT versus other aggregations. dev.to/aws-heroes/u...

@MongoDB drivers do more on the application side than SQL drivers, which mainly send commands as text strings and return simple tabular results. 🍃 BSON is a storage + exchange format that maps documents to app objects. I traced it for a new blog post: dev.to/franckpachot...

The MongoDB text search (in both the Atlas and community editions) uses Lucene indexes. You can check the scoring formulas and parameters with scoreDetails:true dev.to/franckpachot...

🍃 MongoDB transactions are fully ACID, with guarantees across shards and replicas—which is stronger than many SQL databases. Still, myths have a consistent durability 🤷🏼‍♂️ www.linkedin.com/posts/franck...

If you know your application access patterns and cardinalities, you can build a stronger data model with #MongoDB. I explore One-to-Many schemas, queries, and execution plans in a new blog post: dev.to/franckpachot...

Understand how the principle of "store together what is accessed together" is a game-changer for database performance and scalability. By @franckpachot.bsky.social, thanks to @mongodb.bsky.social

Heard of MongoDB index intersection but never seen it? My new blog post explains why. dev.to/franckpachot...

All replicated databases can trade some ACID guarantees for better performance. While not the default, if you perform {w:1} writes on MongoDB, you should be aware of the implications, including local reads and failure scenarios (resilience and failover): dev.to/franckpachot...

Understand how the principle of "store together what is accessed together" is a game-changer for database performance and scalability. By @franckpachot.bsky.social, thanks to @mongodb.bsky.social

A new blog post: MongoDB’s $lookup isn’t like SQL JOINs. It returns related info as arrays in documents, not flat rows. Use $unwind for row-style results, or inner join behavior. And it supports many-to-many with one $lookup on arrays. dev.to/franckpachot...

Meeting many friends at @doagev.bsky.social

My first article on @thenewstack.io is about data locality vs. data independence

Like any database, you can join in #MongoDB, with hash join, nested loop, and indexes. Understanding them and creating indexes (or not) are important for optimal performance. dev.to/franckpachot...

I quickly tested the new query planner in Amazon DocumentDB: dev.to/aws-heroes/a...

Scalable database: query response time stays consistent as data grows. A simple query to compare MongoDB with emulations - typical OLTP with filter, sort, and pagination: dev.to/franckpachot...

I've never realized before that SQL*Plus can tell you when it disconnects "with complications" 🤓

B‑trees face issues like page splits, locks, and buffer pins. LSM‑trees avoid them but raise write amplification. WiredTiger takes a third path: lock‑free B‑trees with hazard pointers for reads and skip lists for writes. Great article on the #MongoDB blog: www.mongodb.com/company/blog...

I'm curious if some @oracleace.bsky.social will be attending AWS re:Invent, given the multi-cloud partnership. Asking because there'll be an event open to advocates, heroes, champions, builders, ambassadors, and user group leaders to celebrate community leaders and meet

Developers — curious to know how you document data model relationships today? www.linkedin.com/posts/franck...

An example of a covering index to avoid fetching large documents in a MongoDB aggregation pipeline dev.to/franckpachot...

I enjoy staying connected with multiple database communities. Even if I share more about one today 🌱🐅, I'm happy to have feedback from others 🅾️🔸️🐘🐬 on what I share

Sad times where tech giant billionaires fire people to build datacenters in a giant game of Monopoly

A migration tool vendor suggests that migration is necessary for improved consistency, reliability, and advanced query capabilities. Yet, success often comes from first understanding your current database. Questioning these claims is the occasion to learn: dev.to/franckpachot...

I'm curious to find out whether Oracle still uses DerbyDB in their Oracle Database Appliances to store the metadata, and what they will replace it with

SELECT DISTINCT ON ... ORDER BY equivalent in MongoDB (and faster) dev.to/franckpachot...

More WiredTiger internals: persisting MongoDB's committed MVCC versions through checkpoints and restarts dev.to/franckpachot...

Trust me I was so against it when I arrived at spark I’ve used it in the past and hated it. but the lack of relations is actually amazing for backfill. it just makes it so simple and painless that I’m willing to live with the mongo of it all

Blog post for September 2025 SQL Developer isn't entirely CBO-friendly. jonathanlewis.wordpress.com/2025/09/24/s...

A closer look at MongoDB Text Search: which terms are indexed for a string, how they're stored, and what formula is used to calculate the query's matching score. Since it's based on Lucene, let's use Lucene tools. dev.to/franckpachot...

Comparing 🍏 to 🍊, and MongoDB to PostgreSQL, with full-text search indexes dev.to/franckpachot...

An example of MongoDB aggregation pipeline compared to the equivalent in SQL/JSON with common table expressions (CTE), from a StackOverflow question: dev.to/franckpachot...

A flexible schema permits arrays in fields without upfront declaration, and filters and indexes (multiple keys per document) use it automatically. MongoDB tracks multikey paths in indexed fields to optimize index scans: dev.to/franckpachot...