- Key/value data store. Couchbase is blazing fast at data access and retrieval when you know what keys you want to access.
- Data querying with N1QL. If you have your indexes set up right, searching your unstructured data can be really fast too.
- Management dashboard. The Couchbase platform has a great admin panel that provides tons of insights into how your cluster is performing.
- The Couchbase mobile suite is great, in theory. In practice we have found the sync to be somewhat unreliable, to the point where we had to write our own logic to push data to the server. We have not had a chance to upgrade to the new 2.0 version of mobile yet, so these issues may have been resolved there.
- 5 years later, the sync gateway component still isn't integrated into the main data platform. The Couchbase platform does a really great job at letting you independently scale services, but the sync gateway is still a standalone component. I get the feeling there is a rift between the Couchbase mobile team and Couchbase server team, or possibly that the Couchbase mobile product is just a side project or second-class citizen in the Couchbase world.
- One of the reasons that initially drew us to Couchbase was their touted dedication to open source development. Over the past year or so, they have really backed off on that. At the 2018 Couchbase connect conference, I didn't even hear the phrase 'open source' mentioned once. If you really dig into the source code that is available, you'll find that pretty much all new features are being developed closed-source in private repositories. So while you can build the 'community edition' yourself, it's not even close to what the full platform offers. So if you do hear any mention of 'Couchbase is open source', be aware that it's a bit of a facade.
If you want to store tons of data records with varying schemas or totally unstructured data, Couchbase is also a great fit. Their N1QL engine is pretty amazing and pretty fast, but don't expect SQL level performance.
If you're building a simple application with traditional relational data and on a tight budget, I'd say stick with something like Postgres.
However, if you need schema validation or enforcement, it lags behind, because that's not the point of the database. Also, because N1QL relies heavily on indexes, I wouldn't recommend it if you plan to do a lot of ad-hoc queries on your production data, as enabling the general index on a production server is not recommended.
It may not be suitable for source of records like member account information that needs to be kept for audit purpose.
- The platform provides the best integration of data delivered via mobile app.
- The platform is good at real-time analysis of multiple data streams.
- The platform is intuitive for admins to manage with good tools for permission setting.
- The administrator functions via mobile are slightly difficult to find.
- Data streams must be correctly integrated as opposed to AI-based integration of the streams.
- For a financial firm, fraud detection is of the utmost importance, and Couchbase does this very well -- but then, you also can leverage this data for better real-time analysis and detection
- We needed strong response times when customer profiles are actively changing and updating, and Couchbase passed with flying colors
- Couchbase does not have a very digestible view for admins, it is more of a secondary thought to maximizing the functionality of the platform
- Couchbase could provide a better UI for updating Server from a mobile or tablet
- Provides ANSI SQL capabilities using N1QL
- Auditing, eventing and analytics in the latest versions of Couchbase
- TTL based document expiry to avoid batch deletes and improve overall cluster performance
- Memory first approach to accessing data
- RBAC and auditing provide excellent security capabilities in limiting privileges and identifying activities
- Consolidate the number of logs and remove generic log messages for quick troubleshooting
- Better error handling capabilities using error# to identify and address critical errors
- Document driven application usage
- Migrating smaller databases from RDBMS to Couchbase
- Less i/o, rather more memory-centric applications
- Horizontally scalable environment
- Data warehouse that requires higher data retention
- Normalized data environments that are structured and are limited to a schema.
- Vertically scalable environments that require higher CPU/memory
- Low latency and high performance
- Easy-to-scale and asynchronous replication
- High scalability
- Live cluster reconfiguration
- High sustained throughput
- Full-text search
- Real-time analytics
- Advanced security and auditing
- Indexing mechanisms need improvement
- JSON support not as good when compared to others
Couchbase Data Platform Review: "Reducing cloud costs and improving performance for our nosql needs"
- Analytics (in the beta version) - being able to run broad queries and derive insights without affecting the operational store (ie. production systems)
- Performance - the memory caching is a big boost from the other nosql dbs we evaluated
- XDCR - super easy for cross region replication
- Needs a managed cloud offering (ie. no worrying about infrastructure, kubernetes or anything.....just provision and go, and no devops required)
- Documentation online could be more consolidated, and more of it :-)
- It's not immediately clear which data to combine into the same bucket, versus which data to put into their own buckets (ie. guidance for which "type"s go together)
- Mobile integration - SDKs that can be written in common languages for mobile development
- Easy failovers - create and duplicate servers with ease
- Direct communication with web app - no need to serialize/convert JSON to relational tables
- Should allow servers to change their roles on the fly - currently you will need to take down the server and re-assign roles
- SDK documentation is not as helpful and should have more examples
- No pre-compiled procedures - unlike in SQL Server, Couchbase does not have pre-compiled stored procedures therefore we have to run various test to optimize our application
- Scalability - As our needs grow we can scale horizontally, taking advantage of autosharding. This is much less cumbersome than dealing with manual partitioning in RDBMS systems.
- Availability - Automatic replication and failover within the cluster means we have no downtime, even if a node fails. It also means that we can scale up or perform database upgrades with zero down time.
- Agility - The schemaless nature of Couchbase allows us to continuously deploy new application versions that add additional data to records, without the overhead and complexity of using DDL to alter table schemas.
- Query Optimization - While the SQL-based query language is very robust and powerful, the automatic query optimization isn't as mature as RDBMS platforms. This requires a bit more attention to indexing and index hints.
- Management UI - The management UI is much better than most other NoSQL platforms, but the need to connect to different nodes in the cluster depending on which function you are performing (FTS queries, N1QL queries, view management) can be a minor frustration.
- Index Replica Management - The process for managing index replicas for high availability is not yet a streamlined as the other services, and can require some manual maintenance.
- Speed and access to data in a very high speed environment
- Scalability to handle large amounts of data
- Reliability and redundancy in a very highly available environment
- Easier upgrade paths to newer versions
We deployed the Couchbase Data Platform to support the new driver statistics service. Couchbase Lite records statistics as the user is driving. Data is synchronized and stored to the Couchbase Server through the Couchbase Sync Gateway. The Coyote service creates aggregated documents, and stores them back in the Couchbase Server. Drivers then retrieve that data through the mobile app or a website.
- Couchbase Data Platform is an End to End solution, from Mobile to Backend. It allows bringing NoSQL databases to embedded devices.
- It is scalable and we can adapt the number of nodes according to our needs (multidimensional).
- The Couchbase support is really good. Each time we needed them, we could have someone on the phone very quickly.
- Our main issue with Couchbase Data Platform is the dependency of the Gateway to MAP / REDUCE views on the server side. Those views may sometimes fail with a high workload and it is the main bottleneck in our current use case.
- Couchbase is currently working on it and we are waiting for the next version which will rely on the N1QL index because they are far more reliable and scalable than views.
Couchbase is also very very good for bringing NoSQL DB to mobile applications including all synchronization process.
For now, it is a little less fitted to handle huge workload from Mobile because of Map/Reduce view dependency.
- The schema-less architecture allow us to provide backward support to system points and make changes at the same time.
- Couchbase lite helps with offline data stores, and makes online syncing very easy.
- Noft a ton of books written about it, most online support is from Couchbase itself.
- Management console can be cumbersome.
- Low latency response time with the help of managed cache at bucket level
- Cross data center replication for high availability across data centers and disaster recovery
- Secondary indexing and N1QL query support
- Real-time scaling and Ease of administration
- Lack of Enterprise features such as centralized administration tool, auditing, replication network.
- Performance, it is way better than Mongo
- Ease of maintanence. Reduced factor in tryng to resolve issues.
- Scales easily and quickly.
- HA across DCs is really needed, which was provided.
- Reduction in pricing
- Slightly better documentation
- Allow community to submit/commit
Couchbase Data Platform Review: "Couchbase Server - full of useful features, despite difficult to navigate documentation"
- Scaling - By adding more hardware to the Couchbase Server cluster, one can quickly benefit from the overall increase in performance of the system
- Mobile Synchronization - As a mobile developer, keeping data in sync between an embedded SQLite database and a RDBMS system takes more effort and thought than necessary and reduces time spent developing the rest of the application. Couchbase Server paired with the Sync Gateway and Couchbase Mobile significantly simplifies those transactions.
- Hands-on documentation via Classes, Seminars, and Tutorials - The free videos on Couchbase's YouTube channel and the courses and seminars that are offered cover much of the basics needed to understand how to quickly get started with Couchbase Server and the features it provides.
- No enforced schema policy - While this is true of most NoSQL technologies, the main benefit that I see Couchbase has is that they have recommended patterns for designing stored data.
- Written documentation - Overall, navigating the documentation site is difficult. There have been several times when I try to search for something and find outdated code references or articles.
- The Couchbase cluster was very stable. We almost never needed any maintenance work for the cluster.
- The throughput was excellent. We started with just a 5-node cluster, and it sustained 700K/sec lookup operations for extended period of time. Later we upgraded to a 20-node cluster, it was able to take 1.2 million read/write operations. That's probably the limit of our client application.
- Very easy on admin side.
- Based on our use case, we don't have anything in particular that's not satisfactory.
Couchbase Data Platform Scorecard Summary
Feature Scorecard Summary
About Couchbase Data Platform
Couchbase is a multi-model NoSQL database for mission-critical applications. It is designed to help deliver ever-richer and ever more personalized customer and employee experiences. Built with powerful NoSQL technology, the Couchbase Data Platform was architected on top of an open source foundation for the massively interactive enterprise. The geo-distributed Engagement Database promises to provide developer agility and manageability, as well as performance at any scale.
Customers include industry leaders Amadeus, AT&T, BD (Becton, Dickinson and Company), Carrefour, Cisco, Comcast, Disney, DreamWorks Animation, eBay, Marriott, Neiman Marcus, Tesco, Tommy Hilfiger, United, Verizon, Wells Fargo, as well as hundreds of other household names. For more information, visit www.couchbase.com.
Couchbase Data Platform Competitors
Couchbase Data Platform Technical Details
|Operating Systems:||Windows, Linux, Mac|