At its GrafanaCONline event, Grafana Labs today announced an update to the open source Grafana dashboard. The update adds visual query tools to make it easier for IT professionals of any skill level to launch queries against the Prometheus monitoring platform or the company’s Grafana Loki log aggregations framework.
In addition, Grafana said the open source edition of Grafana OnCall, a scheduling and paging application, is now also available as open source software that can be used in on-premises IT environments. Previously, Grafana OnCall was only available as part of the managed Grafana Cloud service.
Archana Kesavan, director of product marketing for Grafana Labs, said version 9.0 of the core Grafana platform makes Prometheus monitoring more accessible. Prometheus is being advanced under the auspices of the Cloud Native Computing Foundation (CNCF). It is most widely used in Kubernetes environments, but can also be used with monolithic applications. The latest version of Grafana adds support for an exemplars overlay for traces and Prometheus sparse histograms.
Other capabilities added with this release include a dashboard preview and title search capability, support for envelope encryption by default, an expanded navigation bar and a revamped heatmap panel that renders more quickly than its predecessor.
‘
Finally, the alert system that Grafana Labs first introduced as an option is now a default setting.
Kesavan said in addition to making it easier to launch queries, the Grafana Labs approach also limits the ability of a poorly constructed query to impact clusters running Prometheus.
In general, the adoption of Prometheus has risen alongside Kubernetes. At least initially, DevOps teams that adopt Kubernetes often employ Prometheus to monitor the platform. Less clear is the degree to which IT teams will continue to rely on Prometheus once cloud-native applications are deployed in a production environment. Many IT teams already have a monitoring platform they employ that they choose to extend to Kubernetes environments. Conversely, however, some organizations are opting to replace legacy tools with the Prometheus monitoring platform to monitor both Kubernetes environments and legacy monolithic applications.
At the very least, Prometheus has established a de facto standard for collecting monitoring data from Kubernetes environments that can be queried by a wide variety of tools.
One way or another there will soon be fleets of Kubernetes clusters deployed across the extended enterprise that will need to be monitored. DevOps teams will naturally need to decide for themselves how best to monitor those clusters and, ultimately, observe them in a way that enables queries to be launched and issues surfaced before they become a major problem.
In the meantime, DevOps teams might want to consider consolidating many of the monitoring tools they use today as part of an effort to streamline operations. Rather than having to correlate events across a dozen tools to determine the root cause of an issue, the goal should be to have enough tools to ensure verification without spending hours sorting through conflicting reports.