In my opinion PostgreSQL is totally over MongoDB - not only works with structured data & SQL & strict types, but also has excellent support for unstructured data as separate data type (you can store arbitrary JSONs - and they may be also queryable, depending on one of format's you may choose). Both writes & reads are much faster, then in Mongo. So you can get best on Document NoSQL & SQL in single database..

Formal downside of PostgreSQL is clustering scalability. There's not simple way to build distributed a cluster. However, two points:

1) You will need much more time before you need to actually scale due to PG's efficiency. And if you follow database-per-service pattern, maybe you won't need ever, cause dealing few billion records on single machine is an option for PG.

2) When you need to - you do it in a way you need, including as a part of app's logic (e.g. sharding by key, or PG-based clustering solution with strict model), scalability will be very transparent, much more obvious than Mongo's "cluster just works (but then fails)" replication.

We decided to use MemCachier as our Memcached provider because we were seeing some serious PostgreSQL performance issues with query-heavy pages on the site. We use MemCachier for all Rails caching and pretty aggressively too for the logged out experience (fully cached pages for the most part). We really need to move to Amazon ElastiCache as soon as possible so we can stop paying so much. The only reason we're not moving is because there are some restrictions on the network side due to our main app being hosted on Heroku.

#Caching #RailsCaching

As we've evolved or added additional infrastructure to our stack, we've biased towards managed services. Most new backing stores are Amazon RDS instances now. We do use self-managed PostgreSQL with TimescaleDB for time-series data—this is made HA with the use of Patroni and Consul.

We also use managed Amazon ElastiCache instances instead of spinning up Amazon EC2 instances to run Redis workloads, as well as shifting to Amazon Kinesis instead of Kafka.

We primarily use MariaDB but use PostgreSQL as a part of GitLab , Sentry and @Nextcloud , which (initially) forced us to use it anyways. While this isn't much of a decision – because we didn't have one (ha ha) – we learned to love the perks and advantages of PostgreSQL anyways. PostgreSQLs extension system makes it even more flexible than a lot of the other SQL-based DBs (that only offer stored procedures) and the additional JOIN options, the enhanced role management and the different authentication options came in really handy, when doing manual maintenance on the databases.

One of our battles at the very beginning of the road was choosing the right database. In fact, our first prototype was built on MySQL and back then nothing else was even under a consideration (don't ask me why). At some point, I was working on a project which was running on PostgreSQL and it is only then I understood the full power of it. We have over a billion of records in production instance, and we are able to optimize it to run fast and reliable. Well, now my default DB is PostgreSQL :)

PostgreSQL Heroku Node.js MongoDB Amazon DynamoDB

When I started building Checkly, one of the first things on the agenda was how to actually structure our SaaS database model: think accounts, users, subscriptions etc. Weirdly, there is not a lot of information on this on the "blogopshere" (cringe...). After research and some false starts with MongoDB and Amazon DynamoDB we ended up with PostgreSQL and a schema consisting of just four tables that form the backbone of all generic "Saasy" stuff almost any B2B SaaS bumps into.

In a nutshell:cPostgreSQL Heroku Node.js MongoDB Amazon DynamoDB

When I started building Checkly, one of the first things on the agenda was how to actually structure our SaaS database model: think accounts, users, subscriptions etc. Weirdly, there is not a lot of information on this on the "blogopshere" (cringe...). After research and some false starts with MongoDB and Amazon DynamoDB we ended up with PostgreSQL and a schema consisting of just four tables that form the backbone of all generic "Saasy" stuff almost any B2B SaaS bumps into.

In a nutshell:

• We use Postgres on Heroku.
• We use a "one database, on schema" approach for partitioning customer data.
• We use an accounts, memberships and users table to create a many-to-many relation between users and accounts.
• We completely decouple prices, payments and the exact ingredients for a customer's plan.

All the details including a database schema diagram are in the linked blog post.

We initially chose RethinkDB because of the schema-less document store features, and better durability resilience/story than MongoDB In the end, it didn't work out quite as we expected: there's plenty of scalability issues, it's near impossible to run analytical workloads and small community makes working with Rethink a challenge. We're in process of migrating all our workloads to PostgreSQL and hopefully, we will be able to decommission our RethinkDB deployment soon.

PostgreSQL plus TimescaleDB allow us to concentrate the business effort on how to analyze valuable data instead of manage them on IT side. We are now able to ingest thousand of social shares "managed" data without compromise the scalability of the system or the time query. TimescaleDB is transparent to PostgreSQL , so we continue to use the same SQL syntax without any changes. At the same time, because we need to manage few document objects we dismissed the MongoDB cluster.

Much of our data model is relational, which makes MySQL or PostgreSQL (and family) fit the API's we need to build, in order to meet the needs of our customers.

Sometimes the flexibility of a NoSQL store like Amazon DynamoDB is very useful, but the lack of consistency really impacts usability and performance long-term, compared with viable alternatives. At our current scale, we've seen huge benefits from moving some of our tables out of Dynamo and doing more in SQL.

There will always be use cases for NoSQL and key-values stores, but if your model is well understood in your business/industry: relational databases are the way to go after finding product-market fit. Always understand the trade-offs (and a few intimate details) of any data store before you add to your company's stack!