What needs improvement with Grafana?

Grafana could consider building its own metrics system to eliminate the reliance on other tools like Prometheus, providing a one-stop solution.

The features are complicated and not intuitive. It would be better if they made the technology easy to use without needing to read extensive documentation. The dashboard interface is too wide, and it's challenging to find buttons or features.

Grafana could improve by having its own query language. Currently, it uses languages like Prometheus or InfluxDB, which not everyone knows. Their query language should be easier.

It was pretty complicated to integrate the product into our workflow. We relied on templates we found on the GitHub open-source repositories. It is pretty hard to integrate Grafana and set it up according to our use case. The solution must provide tutorials and guides. We are a small team. We don't have the time to learn Grafana end-to-end, but we want some guides to help us use it quickly. Prebuilt templates will help a lot. The product must provide more alerting metrics. It must be integrated with iOS and Android mobile phones.

I have a problem with Grafana in the area of documentation. There is a problem in the area of some graphs where you connect elements. With Grafana, there is only one name that can be included for ten hours. In Grafana, only one big field with one text name is shown in the documentation. When it comes to production, it is only for a period of time that some fields on each side give the information they have. I researched on Google and other forums and saw that people face problems when they can't connect a field with the same value. The aforementioned area in the tool needs improvement.

The main drawback is the necessity for endpoint monitoring. Due to this requirement, we faced challenges in fetching details related to endpoint monitoring. We are considering the implementation of better tools.

There is a need for improvement in automating daily monitoring reports, especially when alerts are triggered due to system downtimes or fluctuations. This enhancement would make the monitoring tool more efficient and streamlined. While it does produce reports, the quality might not be optimal.

The service dashboard is very hard and needs improvement.

There is room for improvement in terms of stability.

I had issues with the solution's configuration part. Though I resolved it, the configuration is an area needing improvement since it is the solution's only problem.

Setting up alerts via Grafana is a bit complicated, and alerting needs to improve. We used Azure for our alerts and Grafana purely as a dashboard tool. The solution's stability needs improvement.

The documentation or training provided by Grafana is limited compared to its competitors, like Splunk. Some functionalities are unavailable for us since we use the AWS-managed version of Grafana

The solution's interface could be more accessible.

There are not a lot of plugins for financial market monitoring. Our old tool had more plugin capabilities. We need more AI features. We need more data analytics and monitoring done by AI.

They should improve the functioning of the editing tool in the solution's older version.

One thing I got to know is that Grafana doesn't provide anything for reporting. As a DevOps engineer, I would like to get a weekly report expressing the system's overall performance in the last week. For example, how many alerts were triggered, how many APIs responded with bad requests, and how many failures occurred? All these things are generally required at the beginning of the week so that we can better research and optimize everything that's required. Currently, Grafana sends a PDF version of the report, and they are not doing anything beyond that. So, that's a feature that Grafana can work on to make it more accurate and better. In the additional features, I would like to see logs. Logs are basically created when an API is hit. The log contains information such as the API that was hit, the status code, and the time taken. For instance, the log file would store the error message if there was an error in the API. It makes it easier for the developer to debug the API that showed an error. Grafana claims that it has Loki, which is used for logging and saving logs. However, I could not configure Loki with Grafana to see my logs. Grafana could look into this field and make changes to improve it.

I would like to have a better configuration in the next release. The technical support has room for improvement.

There is room for improvement when using multiple dashboards because they can become complicated to keep track of and use. I would like to have alerting functions added to the solution.

We need different kinds of applications in our infrastructure to see information in Grafana. For example, some other companies use OneAgent, which takes control like Grafana but in one place and is better than Grafana.

The solution's integration with other tools needs improvement. I would like the ability to download my results into any format in order to share the information with my clients.

I don't have any notes for improvement. There are some areas of network drives that are not showing as expected based on server usage. On the server, it was showing in GB model. However, in Grafana, It is in bytes. Therefore, the calculation it was not showing correctly. That is something that needs to be changed by the vendor team. The initial setup is not necessarily straightforward.

It is difficult to manage defects because the solution does not provide granular details about the internal system for tracing. Writing queries can be a bit difficult because the syntax must be maintained. The dependency on syntax should be removed so that it does not matter for queries.

The solution has room for improvement with a better API to help automate the construction of the dashboards easier.

In order to enable Grafana, you need another tool called Prometheus, which loads data to it. It would be helpful if Grafana provided more information and training on how to use Prometheus. In the next release, Grafana should include API monitoring.

Grafana has made some changes recently. Previously, we had different dashboards but they have now merged the three designs into one dashboard and now we have to monitor a single dashboard for multiple reasons. It may not seem challenging, but we have critical issues that arise, and we don't know where they're coming from. We have to check all three aspects which takes us longer than it used to and can also affect productivity.

The product could be improved with a feature that provides detailed error messages when we encounter issues. For example, we had a case where there was an error message and we could not decipher what went wrong. Additionally, the product's configuration for saving files could be improved, and more plugins could be provided when using the dashboards.

Perhaps the display of the trigger limits should be more efficient because it's difficult to see that in a graph. We also have an issue with flapping where values move up and down and knowing how to deal with that is quite complicated in Grafana.

The lack of native ability to aggregate data from multiple sources could be improved. As it is not a data store, it is challenging to correlate data from multiple data sources. Adding an out-of-box capability to aggregate and correlate different types of data would make Grafana even better. It is limited on the reporting type supported, which is important for managerial-level officers who want reports that are either general or specific. Improvements in that area would be great. Machine learning capabilities are still limited. This is an improvement in that area is needed. Performance improvement when querying and searching textual data would be beneficial.