#314 — July 24, 2020

Read on the Web

This is a pretty good issue, even if we do say so ourselves 😄 Quite a few interesting releases, plus Jaana's Spanner story, and a neat database 'tube map' from yesteryear make this worth more of a skim if you've skipped any issues recently 😉

Database Weekly

Analyzing 1.1 Billion Taxi Rides using OmniSciDB on a MacBook — This is the latest in a line of experiments Mark has run to analyze a large amount of data with different database systems. This time it’s the turn of OmniSciDB (formerly known as MapD), a SQL-based columnar database engine focused on performance via parallelization. Spoiler? “To get this level of performance on a regular piece of office equipment is a big game changer.”

Mark Litwintschik

Spanner’s SQL Story — Spanner is a distributed ‘NewSQL’ database Google built with global distribution in mind (such as for transactions and replication). There has been some confusion over its actual SQL support, though, and Jaana clears up Spanner’s SQL story here.

Jaana Dogan (Google)

Free Download: Kubernetes Best Practices (O’Reilly) — In this new manual, four pros (including Kubernetes cofounder Brendan Burns) guide you through best practices of building applications on Kubernetes.

Cockroach Labs sponsor

Introducing Apache Cassandra 4.0 Beta — Cassandra, a wide column NoSQL database, has inspired numerous other systems including ScyllaDB, and it’s back boasting “the most stable” release: 4.0. It’s still in beta, but clearly a huge effort has been placed on stress testing, monitoring and observability.

Apache Cassandra Team

⚡️ Quick bytes:

  • Nebula Graph, a distributed graph database system and company, has reached 1.0 and received $8m in funding. We hope to see more from these folks  soon.
  • New memory and compute optimized hardware options are now GA on Azure SQL Database. How does 128 vcores and 4TB of memory for your database sound? (Expensive, I imagine.)
  • An ongoing attack called 'Meow' is wiping unsecured data all over the place but with no obvious reason why. Fingers crossed they're doing us a favor.

RavenDB 5.0 Released: An ACID NoSQL Document Database — RavenDB is a NoSQL Database that’s fully transactional and scales to 1 million reads and 150,000 writes per second. There’s no release blog post yet, but we’re told 5.0 just officially came out after some last minute stress testing snafus which are now all resolved yielding a truly ‘monkey resistant’ build :-) A big part of the 5.0 release is time series data support.


Demystifying Database Systems: Isolation Levels vs. Consistency Levels — Sometimes the concepts of isolation and consistency can get mixed up, especially when a vendor is trying to push the benefits of one or the other.

Daniel J. Abadi

Leaving Oracle for PostgreSQL? You’re Not Alone — We’ll help your developers migrate to PostgreSQL making sure application code transitions to Oracle-compatible functions.

EDB sponsor

How Shopify are Solving Their Data Discovery Challenges — As one of the biggest providers of ecommerce services, Shopify has a lot of commercially valuable data to deal with, and this post introduces us to Artifact, a tool they’ve built for data accessibility and governance purposes. It’s proprietary (for now) but may give you some ideas.

Shopify Engineering

Why ListenBrainz Moved from InfluxDB to TimescaleDB — In short, they’re more familiar with Postgres overall and didn’t want to keep a separate set of mental models in mind to stay with Influx.


Beyond Relational Databases: A Focus on Redis, MongoDB, and ClickHouse — A rather high level white paper that you have to enter your name and email to download, but.. this landing page is worth checking out for the neat public transit style map of database systems too 😁


FastoNoSQL: A Cross-Platform Redis, Memcached, SSDB, LevelDB, RocksDB, UnQLite, LMDB, etc. Client — If you’re using one of numerous non-SQL databases, this cross-platform, native (built using Qt and C++) client may be of interest to you.