Manager, Software Engineering at a tech vendor with 10,001+ employees
Real User
Top 5
Aug 19, 2025
We have been using CockroachDB for the last two years. From 2023, we started working on a product similar to Google Capture. We decided to work with CockroachDB with Hasura. Hasura is a UI for the database CockroachDB that provides a playground to work with the API and data. Since CockroachDB primarily supports SQL, we didn't want to use plain SQL in our programming, so we decided to go with ORM and selected Hasura. With Hasura, we sometimes require GraphQL. Besides using GraphQL, we have also used SQL with CockroachDB directly to get data and perform insert and update operations. Regarding our operations, CockroachDB equals PostgreSQL plus Google Spanner, and we can remove scalability, then add resilience across regions. For CockroachDB multi-active availability, all replicas of our data are active and can serve read and write simultaneously. Currently, we are not using these features because our database is not too large with only 20 customers using our application. MAA is a feature provided by CockroachDB, but they charge for this functionality. It has consequences by raft where each range of data is replicated with default three replicas. Writes go through raft consensus where a majority of replicas must agree, ensuring strong consistency even if some nodes or regions fail. Regarding multi-region or global clusters, it allows spanning clusters across multiple regions. We are configured with AWS because we use Lambda functions on the server end. The zone configuration lets you control where replicas live, such as keeping EU customer data in EU and US customer data in US, though we aren't currently using this feature. ACID transactions are not required in our application because we don't have any payment gateway currently. CockroachDB uses a two-phase commit protocol on top of raft consensus. If any part of the transaction fails, commits, times out, or node crashes, CockroachDB automatically rolls back. The benefits include being developer-friendly, strong security with RBAC, encryption, and data residency laws by pinning data to specific regions. It is self-healing because if a node crashes, data automatically rebalances to healthy nodes without manual failover, replication, or clustering needed. Finally, it offers cost efficiency.
I am studying how to deploy CockroachDB and YugaByteDB, and learning some basic information about them. I am testing these databases as part of my school application to find a suitable database for our applications. Currently, I am using PostgreSQL, however, I want to try some distributed databases for testing purposes.
Software Engineer at a consultancy with self employed
User
Nov 15, 2023
Cockroach nodes were installed in the following: 1. Single host, triple nodes (containers): for evaluation on a low-end PC. 2. Single host, triple nodes (process): to test applications against a ~500GB database in-house. 3. Serverless: hosted in Google Cloud Platform (the main database). A number of Python scripts and some Java applications are happily reading and writing to the database. The solution allows for scaling in cases where a PostgreSQL server (unless you use sophisticated partitioning across some machines) would not be enough to handle the load. This kind of database is particularly used for backtests.
Master Student at a university with 11-50 employees
Real User
May 27, 2022
I use CockroachDB to test big data samples and to create the best structure for databases. We have four users and required 10 people for deployment and maintenance.
Oracle Cloud Infra Architect at Sterlite Technologies Ltd
Real User
May 25, 2020
We are currently evaluating this product to see if it can support distributed transactions with high availability. We are comparing it with Oracle, trying to see it can achieve the same thing.
Cockroach Labs is the creator of CockroachDB, the cloud-native, resilient, distributed SQL database enterprises worldwide trust to run mission-critical AI and other applications that scale fast, avert and survive disaster, and thrive everywhere. It runs on the Big 3 clouds, on prem, and in hybrid configurations powering Fortune 500, Forbes Global 2000, and Inc. 5000 brands, and game-changing innovators, including OpenAI, CoreWeave, Adobe, Netflix, Booking.com, DoorDash, FanDuel, Cisco,...
We have been using CockroachDB for the last two years. From 2023, we started working on a product similar to Google Capture. We decided to work with CockroachDB with Hasura. Hasura is a UI for the database CockroachDB that provides a playground to work with the API and data. Since CockroachDB primarily supports SQL, we didn't want to use plain SQL in our programming, so we decided to go with ORM and selected Hasura. With Hasura, we sometimes require GraphQL. Besides using GraphQL, we have also used SQL with CockroachDB directly to get data and perform insert and update operations. Regarding our operations, CockroachDB equals PostgreSQL plus Google Spanner, and we can remove scalability, then add resilience across regions. For CockroachDB multi-active availability, all replicas of our data are active and can serve read and write simultaneously. Currently, we are not using these features because our database is not too large with only 20 customers using our application. MAA is a feature provided by CockroachDB, but they charge for this functionality. It has consequences by raft where each range of data is replicated with default three replicas. Writes go through raft consensus where a majority of replicas must agree, ensuring strong consistency even if some nodes or regions fail. Regarding multi-region or global clusters, it allows spanning clusters across multiple regions. We are configured with AWS because we use Lambda functions on the server end. The zone configuration lets you control where replicas live, such as keeping EU customer data in EU and US customer data in US, though we aren't currently using this feature. ACID transactions are not required in our application because we don't have any payment gateway currently. CockroachDB uses a two-phase commit protocol on top of raft consensus. If any part of the transaction fails, commits, times out, or node crashes, CockroachDB automatically rolls back. The benefits include being developer-friendly, strong security with RBAC, encryption, and data residency laws by pinning data to specific regions. It is self-healing because if a node crashes, data automatically rebalances to healthy nodes without manual failover, replication, or clustering needed. Finally, it offers cost efficiency.
I am studying how to deploy CockroachDB and YugaByteDB, and learning some basic information about them. I am testing these databases as part of my school application to find a suitable database for our applications. Currently, I am using PostgreSQL, however, I want to try some distributed databases for testing purposes.
Cockroach nodes were installed in the following: 1. Single host, triple nodes (containers): for evaluation on a low-end PC. 2. Single host, triple nodes (process): to test applications against a ~500GB database in-house. 3. Serverless: hosted in Google Cloud Platform (the main database). A number of Python scripts and some Java applications are happily reading and writing to the database. The solution allows for scaling in cases where a PostgreSQL server (unless you use sophisticated partitioning across some machines) would not be enough to handle the load. This kind of database is particularly used for backtests.
The solution is used for Kubernetes to put on database.
I use CockroachDB to test big data samples and to create the best structure for databases. We have four users and required 10 people for deployment and maintenance.
We are currently evaluating this product to see if it can support distributed transactions with high availability. We are comparing it with Oracle, trying to see it can achieve the same thing.
The use case is a backend DB for an app. The CockrorchDB is on CentOS7.x with a cloud server service.