InfluxDB Cloud Serverless

InfluxDB  is the main component in our large enterprise-scale streaming data application for maritime vessels. We collect position data from vessels around the coast once per second, put it on a Kafka stream, and feed those positions into InfluxDB  continuously. This has been working flawlessly since 2018. We have seven years of time-series data for all the vessels that my company operates, roughly 130 to 140 vessels. Every  move they make is being tracked and stored in InfluxDB.

We mainly write and read data from InfluxDB. We perform very simple queries to do time series on a key, which is a unique ID of the vessel. We will select a vessel and select from time to time stamp. That’s what we do. InfluxDB’s core functionality is crucial as it allows us to store our data and execute queries with excellent response times.

It is challenging to get long-running backups while running InfluxDB in a Microsoft Azure  Kubernetes  cluster. Replicating data for on-prem development and testing is difficult. Having a SQL abstraction in InfluxDB could be beneficial, making it more accessible for teams that prefer querying with SQL-style syntax.

In total, I've been using InfluxDB for six years.

InfluxDB is extremely stable. It serves as the backbone of our application, and its stability is crucial. If InfluxDB stops or doesn’t scale, the entire application stops.

Scalability is critical. We’ve scaled on volume with seven years of continuous data without performance degradation. The scalability allows us to track vessel movements per second back to the application's conception in 2017.

We did not use a different solution for these use cases before InfluxDB.

The initial setup can be intimidating for newcomers, and there is a certain threshold needed due to the performance we get. However, once familiar with the setup, it becomes streamlined.

Four people were involved in the deployment process, and one person is now needed for maintenance.

We haven't gauged any measurable benefits; our company is more operational-focused.

We use the open-source version of InfluxDB, so it is free.

We evaluated using PostgreSQL  and a time-series database in Amazon, though I can't recall its name.

InfluxDB works as expected with excellent scalability and stability, which is critical for our application. I rate InfluxDB ten out of ten overall.

Hybrid Cloud

We are developing a trading agent that uses multiple machine learning models to adapt to the crypto market in real time. InfluxDB  is used to collect data on crypto coin prices from exchanges like Binance and Bybit. Our use case requires low latency and the ability to query data effectively. We use InfluxDB  on a DigitalOcean  infrastructure in a containerized environment with Docker .

The most important feature for us is low latency, which is crucial in building a high-performance engine for day trading. InfluxDB can handle around ten thousand messages per second, which is essential for our requirements. The solution's ability to store time series data is also significant in our crypto trading use case where time series data about prices is critical.

One area for improvement is the querying language. InfluxDB deprecated FluxQL, which was intuitive since developers are already familiar with standard querying. Though we can adapt to the Flux  language, I would like to see more development in this area and am unsure why FluxQL was deprecated.

We have been using InfluxDB for the last eight months.

We did not encounter any issues with the deployment. Using Kubernetes  allowed us to easily set up InfluxDB in a containerized environment. Although DigitalOcean  does not offer a managed database service, deploying our own container was straightforward and aligned with our continuous integration processes.

We have not experienced any stability issues with InfluxDB so far, and it has been acceptable for our needs.

Scalability has not been an issue because we have only used one instance of InfluxDB. It is primarily used for real-time data acquisition rather than for extensive scaling.

We have not needed to contact technical support. All resources required were available through documentation, enabling us to resolve any issues on our own.

Neutral

Previously, we used CassandraDB and ScyllaDB , a fork of CassandraDB. While these were performant, they did not store data in the time series format essential for our needs. Once we discovered that there were databases like InfluxDB designed for time series data, we decided to try it.

The initial setup was straightforward, as we used Kubernetes  to deploy InfluxDB. Although DigitalOcean does not offer a managed database service for InfluxDB, setting up our own container was an easy process.

One person was responsible for the entire deployment of InfluxDB in our organization.

I have experience with CassandraDB and ScyllaDB  as alternatives.

My advice for new users would be to ensure you are choosing the right engine for your domain. For InfluxDB, it performs well for low latency inputs and high-performance real-time data. While I would rate InfluxDB a ten on a scale of one to ten, users should be thoughtful about matching the engine to their specific needs.

Public Cloud

Other

I use the solution to store and manage data from various sensors in a production environment. I have developed a system where data from these sensors is communicated through an OPC UA receiver and stored in InfluxDB. It handles serial data from sensors effectively and integrates well with Grafana for visualization.

The platform operates very quickly. It is easy to configure, connect, and query and integrates seamlessly with Grafana.

I have been using InfluxDB for four years.

I have not experienced any stability issues with the product. 

Approximately 20% of our team uses InfluxDB.

I recommend InfluxDB to others and rate it a ten. 

What do you like best about the product?

InfluxDB allows us to store large quantities of data in real-time that originate from our telemetry systems. Additionally, its simplicity and speed in performing backups using the provided tools ensure robust data security without additional complications.

What do you dislike about the product?

What I dislike about InfluxDB is the limited support and updates for older versions like 1.8, which are still widely used.

What problems is the product solving and how is that benefiting you?

InfluxDB solved our problem of storing the telemetry data from our prototypes and make it accesible to the rest of team. Also, allows us to querty the data in real time, simplifing our debug process.