VictoriaMetrics Cloud: Prometheus & OpenTelemetry metrics and logs

VictoriaMetrics  is a time-series database that we use to handle a large volume of metrics coming from our application. We have multiple applications running, and we want to monitor everything to understand exactly what is happening, so we deployed VictoriaMetrics  to serve as our time-series database and then use Grafana  to visualize the data. VictoriaMetrics serves as a replacement for Prometheus to monitor all the time-series data that comes from our applications.

VictoriaMetrics is built with memory constraints in mind, and we saved a significant amount of memory after implementing it. The first advantage is that it is quite easy to use, and we can leverage PromQL queries in VictoriaMetrics with a straightforward deployment process.

The best aspect of using VictoriaMetrics is the reduction in resources including RAM, memory, disk, and other infrastructure requirements, which have all decreased substantially after implementation.

We observed that RAM was reduced by around 70% and disk usage by 75%, while write speed increased 3x and query per second P95 latency became 7x faster. The financial savings are significant because we observed a substantial reduction in resources such as CPU and memory after adopting VictoriaMetrics.

VictoriaMetrics is quite a good project and quite easy to use overall, though some documentation could be improved. The main area that needs improvement in VictoriaMetrics is the documentation itself, as the support is adequate and the open-source project does provide replies within one to two days.

We have been using VictoriaMetrics for around two years.

VictoriaMetrics is stable.

VictoriaMetrics's scalability is quite good.

We use the open-source version of VictoriaMetrics, so we did not require any customer support. However, we do use the open-source project to contact them and learn more about the issues and projects that we are working on.

We used Prometheus earlier, but we found that the resource usage was quite high, so we decided to try VictoriaMetrics, which resulted in a substantial reduction in resources.

We did not require any license for VictoriaMetrics since we are using the open-source version, so the setup cost is what we incurred.

We did not require any license for VictoriaMetrics since we are using the open-source version, so the setup cost is what we incurred.

Prometheus was the solution we were using, and we did not evaluate other options; we simply proceeded with VictoriaMetrics and have continued to use it.

You should use VictoriaMetrics if you are currently using Prometheus, as I believe it is quite helpful and you should consider switching. I give this review an overall rating of 9.

My main use case is storing and querying the time-series metrics for monitoring and observability. I primarily use it as a high-performance back end for Prometheus, where it handles large volumes of metrics data efficiently. In my day-to-day workflow, application and infrastructure metrics are scraped via Prometheus and stored in VictoriaMetrics . I then use it with visualization tools like Grafana  to monitor system health, track performance, and troubleshoot issues. This setup helps me handle high data ingestion with lower resource usage.

One more thing I would add is how it helps with scalability and long-term data retention. I use it to store metrics over long periods without a significant increase in storage cost, which is very useful for trend analysis and capacity planning. It allows me to look back at historical data and make better decisions about scaling and performance optimization. Also, its ability to handle high ingestion rates with consistent performance makes it reliable for production environments, especially when monitoring multiple services and infrastructure components at scale.

When it comes to the best features, in my experience, performance, efficiency, and scalability are key aspects. High performance and fast querying are one of the biggest strengths. It is very fast for data ingestion and query performance. Efficient storage and compression are significant advantages, where VictoriaMetrics  uses strong data compression, allowing it to store significantly more data using less disk space, sometimes up to ten times more efficient than other solutions. Scalability is also a major strength, as it scales very well both vertically and horizontally, making it suitable from small setups to large production environments. Additionally, seamless Prometheus compatibility and low resource usage are valuable.

The efficient storage and compression in VictoriaMetrics has a direct impact on both cost and performance in my workflow. Since it stores metrics in a highly compressed format, I am able to retain longer periods of data without needing additional storage. It also improves performance during queries, even with large volumes of historical data. Queries remain fast, which helps in quick troubleshooting and dashboard loading. I do not have to worry about deleting older data aggressively, so my teams can perform better trend analysis and capacity planning. Overall, it gives me the scalability to monitor effectively.

The feature I really appreciate is flexible query capabilities, MetricsQL. It extends Prometheus's query language and allows more powerful and efficient queries, especially when dealing with large data sets or complex monitoring scenarios. Also, its high availability and reliability are strong points. Even under heavy load, it maintains consistent performance without frequent tuning, which is important for production monitoring.

I have noticed specific outcomes. The biggest improvement is better performance and stability. I am able to handle a much higher volume of metrics without performance issues, which makes my monitoring more reliable in production. I have also seen significant cost optimization, mainly due to its efficient storage and lower resource usage compared to my earlier setup. I need less CPU, memory, and disk. Another key impact is faster troubleshooting and visibility.

I have achieved around fifty to sixty percent reduction in storage usage due to its compression. In terms of performance, query response time improved by thirty to forty percent. I also saw a fifty to sixty percent reduction in CPU and memory usage compared to the previous setup. Additionally, I am now able to retain metrics for a much longer duration without increasing storage significantly.

Overall, VictoriaMetrics is a very strong tool. One area for improvement is documentation and learning resources. While the documentation is good, some advanced use cases and configuration could be explained more clearly with real-world examples. Another improvement would be the UI and built-in visualization. VictoriaMetrics mainly relies on tools like Grafana , so having a more feature-rich native UI for basic monitoring and exploration would be helpful.

On the integration side, while it works great with Prometheus and Grafana, expanding native integration with more tools like CI/CD platforms, alerting systems, and cloud-native services would be beneficial. In terms of support and troubleshooting, having more built-in diagnostics or guided debugging tools would be helpful. Right now, when issues happen, it often relies on logs and external tools. Also, for enterprise usage, strong features like RBAC, audit logs, and multi-tenant management would be enhanced.

The improvements would include UI and built-in visualization. Also, query complexity can be a bit challenging for new users. Simplifying  it or providing better guidance for query building would improve usability. Finally, enterprise features and ecosystem integration could be expanded further.

I have been working with VictoriaMetrics for the past two years.

VictoriaMetrics is very stable.

Scalability is very good. That is why I am getting the points of the performance improvements around thirty to forty percent improvement. The faster queries and dashboards make quicker troubleshooting, which reduces the downtime and improves the system reliability as well. Overall, the scalability is very good.

The customer support is very good. They are following a ticketing system. Whenever I raise any ticket, the customer support team will actively reach out to me and they are helping in a nice manner.

This is my first solution in this category.

I have cost savings due to its efficient compression and lower resource usage. I have reduced storage costs by fifty to sixty percent. I also gained ROI through performance improvements, around thirty to forty percent. Another key factor is operational efficiency with lower CPU and memory usage, around forty to fifty percent reduction.

The pricing was very good as I am following a subscription-based model. As of now, I have a better dashboard to view my monthly and yearly bills. It is very helpful to reallocate my budget to accommodate according to my organization's requirements. Overall, the pricing is very good.

VictoriaMetrics is an excellent solution for time-series monitoring with outstanding performance. The only reason I rate it a nine out of ten is due to areas like document clarity, query complexity for new users, and having limited native UI features. Overall, it is a highly reliable and cost-effective tool for large-scale observability. I would recommend VictoriaMetrics because it is very helpful in performance improvement. I recommend integrating it with your local dashboards, whether you are using Grafana or a DataDog dashboard or something similar. Overall, it is a very good choice. I give this product a rating of nine.