July 2021 Summaries

Jacob Colvin, a Site Reliability Engineer, shares a creative project that uses Grafana and Prometheus to recreate the Rickroll meme as a video rendered through these platforms. The project leverages recent advancements like Prometheus's metric backfilling capability and Grafana's color overrides to animate videos by mapping video frames onto Prometheus metrics and using Grafana for color visualization. With tools like Go's image/png package and grafana-image-renderer, Colvin was able to automate the process of converting video frames into Prometheus metrics and rendering them into a playable format in Grafana. Although the project faced challenges such as handling color complexities and maintaining audio synchronization, solutions like using a screen-ception technique for color and adjusting frequencies for audio were implemented. The culmination of this project was showcased in a lengthy Prometheus Dev Summit video, highlighting the community's contributions and the technical feats achieved through these platforms.

Grafana Labs has been collaborating with Elastic to enhance the Elasticsearch plugin for Grafana, aiming to support data sources favored by users, such as Elasticsearch. This partnership began even before its official announcement, with both teams working together on a shared roadmap to improve the plugin's reliability and performance. Key developments include the complete rewrite of the Elasticsearch query editor using React and the addition of features in Grafana 8.1, such as support for searching frozen indices and rate aggregation, allowing for more efficient data analysis. The collaboration has also led to various fixes and improvements, with future plans focusing on enhancing user experience through features like syntax highlighting, SQL support, and advanced observability use cases. The teams are committed to continuing their partnership to better integrate Elasticsearch with Grafana, benefiting from contributions and feedback from key individuals in the project.

Grafana Labs, known for being a significant contributor to Cloud Native Computing Foundation (CNCF) projects such as Prometheus, Cortex, Thanos, Jaeger, and OpenTelemetry, has elevated its commitment to the open-source community by joining the CNCF Governing Board as a Platinum member. The company's open-source projects, including Grafana, Grafana Loki, and Grafana Tempo, are integral to the cloud-native ecosystem, with a CNCF survey revealing that a significant number of respondents use Prometheus and Grafana together. Since joining CNCF as a Silver member in 2017, Grafana Labs has maintained a robust presence at KubeCon + CloudNativeCon events. The company's increased involvement is further highlighted by Tom Wilkie, VP of Product and a prominent figure in Prometheus and Cortex, who will represent Grafana Labs on the Governing Board, reflecting the company's commitment to leadership and innovation within the open-source community.

Grafana offers powerful integrations with popular software development tools like GitHub, GitLab, Jira, and ServiceNow, providing enhanced visibility into software operations through its data source plugins, particularly beneficial for users of Grafana Cloud or those with a Grafana Enterprise license. In a webinar hosted by Grafana Labs Senior Engineering Manager Justin Hunthrop, viewers can learn best practices for using these plugins to gain insights into development team performance without micromanaging, creating a comprehensive view of software operations metrics. The webinar also covers additional Grafana Enterprise data source plugins, and users are encouraged to explore further resources on the Solutions page or contact the Grafana team for more information. Grafana Cloud offers a free 14-day trial of its Pro version, which includes unlimited metrics, logs, traces, long-term retention, and access to one Enterprise plugin, making it an accessible entry point for those new to observability.

Grafana Cloud has introduced a new integration for Apache Kafka, offering comprehensive monitoring capabilities through a series of dashboards that provide insights into various Kafka components, such as Broker clusters, Zookeeper clusters, Kafka Connect clusters, Schema Registry clusters, and ksqldb clusters. These dashboards monitor key metrics including broker health, topic throughput, consumption lag, and task statistics, utilizing data collected primarily via JMX exporters and an external exporter embedded in the Grafana Agent. This integration aims to enhance observability by bringing together metrics, logs, and traces within Grafana's composable platform, making it easier for users to monitor and manage their Kafka ecosystems. Grafana Cloud offers both free and paid plans, encouraging users to explore its capabilities for observing and alerting on Kafka and other supported technologies, with additional resources and community support available.

Grafana provides a comprehensive solution for managing service level objectives (SLOs) by integrating multiple monitoring data sources into a unified platform, facilitating the visualization and management of service level indicators (SLIs) and error budgets. Grafana Enterprise and Grafana Cloud offer capabilities for combining disparate data into interactive dashboards, allowing organizations to monitor and act on potential SLO breaches efficiently. These platforms support the creation and management of SLIs through PromQL and APIs, with integration possibilities for alerting systems like PagerDuty. This functionality is crucial for site reliability engineers (SREs) to maintain application health and performance without overwhelming teams. Grafana's versatility and scalability enable monitoring of numerous applications and SLIs, ensuring system uptime and reliability.

Grafana Labs has acquired Pace.dev, a London-based startup founded by Mat Ryer and David Hernandez, to enhance its developer experience and expand its offerings in software development tools. Mat and David, known for their minimalist design and expertise in creating developer tools, previously founded Machine Box, bringing machine learning capabilities to developers through Docker containers. The acquisition is rooted in a shared vision of improving technology and strong connections to open-source projects and tech communities. Grafana Labs looks forward to collaborating with the Pace.dev team to build innovative solutions for engineering teams.

Grafana 8.0 introduces a new unified alerting system that consolidates Prometheus and Grafana alerting into a single interface, enhancing the user experience for all Grafana users across different data sources, whether they utilize the open-source or Enterprise and Cloud versions. This update addresses a popular feature request, offering a streamlined process for viewing and editing alerts. To elaborate on this new system, a webinar titled "Intro to unified alerting in Grafana" is scheduled, where Engineering Director Dee Kitchen and Senior Product Manager Richard Lam will guide participants through creating and managing alerts in the updated system. This session aims to help users transition from traditional Grafana dashboard alerts and understand the improvements made, while also demonstrating how the new system simplifies managing Prometheus-style and Grafana Cloud alerts directly within Grafana.

Pernod Ricard, a leading company in the wine and spirits industry, has successfully scaled its direct-to-consumer e-commerce operations using advanced monitoring tools like Grafana and Loki. As the demand for online sales surged, particularly during the COVID-19 pandemic, the company needed to create a scalable, consumer-centric platform that could be rapidly deployed by its global affiliates. Utilizing Magento, AWS, and custom in-house modules, Pernod Ricard established an e-commerce framework known as the E-commerce Factory, which allows for quick deployment and customization of websites. The infrastructure is monitored using a stack that includes Cloudwatch, Container Insights, and Loki, with centralized data accessible through Grafana dashboards. This setup enables developers to independently manage and troubleshoot issues, ensuring a consistent and high-quality customer experience across all affiliate sites. The implementation of Grafana and Loki has harmonized metrics and improved infrastructure scalability, supporting the company's growth while fostering an open-source mindset that Pernod Ricard values.

During a hackathon at Grafana Labs, Richard "RichiH" Hartmann invested time in updating Kraken, a project for reading sensor data, to support modern hardware and libraries. The blog post focuses on using the Bosch BME 680 sensor, which measures air temperature, humidity, pressure, and air quality, as a beginner-friendly project for monitoring environments such as home offices or cellars. The process involves using Grafana Cloud, a managed observability platform, though it is also compatible with local Prometheus setups. Hartmann provides a detailed guide on assembling the necessary components, such as the ESP32 development board and the Bosch BME sensors, and coding them using the Arduino IDE to send data to Grafana Cloud. The article emphasizes the importance of using safe soldering practices and offers advice on setting up dashboards and alerts in Grafana Cloud to analyze the collected data. Additionally, Grafana Labs is actively hiring and encourages those interested in participating in open-ended hackathons to explore career opportunities with them.

Grafana offers powerful tools for visualizing business performance, particularly through the use of cohort tables, which help analyze customer behavior over time. Utilizing data sources such as Google BigQuery, Grafana enables users to understand business health by tracking key metrics like conversion rates and user engagement. The process involves collecting anonymous pageview data with a frontend tracking script, routing it through RudderStack to BigQuery, and then using SQL queries in Grafana to create cohort tables that can be visually enhanced with threshold-based colors to highlight successful outcomes. This method allows businesses to assess the impact of initiatives and improve user experiences, as demonstrated by the increased engagement during GrafanaCONline 2021. By following these steps, organizations can harness Grafana to deliver valuable business intelligence insights.

Synthetic monitoring, as discussed by Yashvardhan Kukreja at GrafanaCONline 2021, is a method to simulate user behavior for testing application performance from a user's perspective, offering valuable insights into large-scale applications. Grafana Cloud's synthetic monitoring plugin is highlighted for its ease of setup and cost-effectiveness, as it requires no coding and allows users to pay only for stored metrics. This approach facilitates the monitoring of End-to-End availability, service-level availability, geographical bottlenecks, and infrastructure-level performance, making it particularly crucial for applications operating across multiple regions. The Grafana Cloud solution provides an automatic dashboard for visualizing synthetic monitoring checks, enhancing the ability to define service-level indicators and agreements around availability and performance. Kukreja emphasizes the plugin's simplicity and efficiency, demonstrated through a live demo, and the session recordings are available on-demand for further learning.

The new Jira integration for Grafana Cloud enhances the capability of organizations to monitor and visualize their software development processes by bringing deeper insights into their Jira Data Center or Server instances. This integration is particularly beneficial for tracking key metrics such as license expiration, user slot availability, and server latency. Although it does not support Jira Cloud instances, it complements the existing Jira Enterprise plugin by offering preconfigured dashboards and alerts that facilitate quick setup and monitoring. These include alerts for license expiration, user account capacity, and email failures, enabling users to maintain and optimize their Jira environments effectively. Grafana Cloud offers flexible subscription plans, including free and paid options, to help users get started with observing metrics, logs, traces, and dashboards, thereby reinforcing its utility in streamlining project management and software development workflows.

Load testing is crucial for assessing the performance and scalability of SQL databases, especially under high user volumes, which can cause bottlenecks. The process involves simulating realistic user flows and testing infrastructure components, such as databases, in isolation to gain insights into their capabilities and limitations. Grafana k6, a popular tool for performance testing, offers extensions like xk6-sql to support custom protocols and facilitate database load testing. This allows users to plan database architecture and scaling more efficiently. A practical example using k6 involves building a custom k6 binary with SQL extensions, writing JavaScript scripts for database operations like setup, execution, and teardown, and running tests with varying numbers of virtual users to determine the database's handling capacity. The tutorial emphasizes the ease of testing databases separately and invites users to explore k6's community and resources, including Grafana Cloud, which offers a free tier and tailored plans for diverse performance testing needs.

The SOFIA airborne observatory, a collaboration between NASA and the German Aerospace Center, uses a modified Boeing 747SP with a 2.7-meter reflecting telescope to conduct astronomical observations above the Earth's atmosphere, offering flexibility similar to satellites but with the ability to return home and swap instruments. During GrafanaCONline 2021, Ronan Higgins from the University of Cologne detailed how Grafana, integrated with the upGREAT instrument software stack, plays a crucial role in monitoring SOFIA's subsystems by consolidating diverse logging and graphing systems into a unified platform. The team channels around 2GB of data per flight into InfluxDB and Grafana, enabling real-time data visualization and quality assessment. Grafana dashboards allow astronomers to identify and address data quality issues, such as noise performance, during flights by correlating these problems with aircraft conditions. Looking forward, the team plans to utilize Grafana's alerting system more extensively for in-flight data quality alerts and aims to develop a custom plugin for linking data points to astronomical data. The adoption of Grafana in astronomy is expanding, with a trend towards system-wide monitoring using this tool, which has significantly improved data handling on SOFIA.

Grafana Labs utilizes Tanka for deploying workloads to Kubernetes clusters, and as their organization expanded, they faced challenges in managing workload configurations consistently at scale. Initially, configurations were manually applied, leading to issues like configuration drift and increased engineering burdens due to the growing number of clusters and environments. To address these, Grafana implemented Continuous Deployment, reducing some issues but not eliminating the complexity of managing numerous Tanka environments. They adopted Jsonnet with Tanka inline environments, allowing for dynamic environment creation using Jsonnet's capabilities, thus mitigating issues like configuration drift and bootstrapping new environments. By integrating the tanka-util library, they streamlined the configuration process, making it easier to manage different cluster environments while accommodating cluster-specific needs. They also automated namespace creation processes, ensuring consistency across environments and reducing manual interventions. This approach views the Jsonnet codebase as a "database," which centralizes valuable information, allowing for reuse and reducing data duplication across different scenarios.

Grafana Enterprise Metrics (GEM) 1.4 introduces significant features such as cross-cluster query federation and out-of-the-box self-monitoring to enhance the multi-tenanted, horizontally scalable Prometheus-as-a-Service experience. The cross-cluster query federation feature allows users to query multiple GEM clusters simultaneously by using a federation frontend component that aggregates responses from various targets, including Prometheus-compatible APIs like Thanos and Cortex, facilitating hybrid metrics setups. This feature is beneficial for handling scenarios like maintaining separate production and development clusters, ensuring data privacy by geographical restrictions, and providing redundancy. Meanwhile, the self-monitoring feature is designed to be deployable with minimal setup across various architectures, offering automatic health metrics visualization through a dedicated system tenant and pre-configured dashboards. These enhancements aim to provide centralized observability teams with robust tools for maintaining and monitoring their metrics infrastructure efficiently.

Salesforce, the leading customer relationship management platform, utilizes Grafana and Prometheus to effectively manage service health and alerts, ensuring product availability insights across the company. During a GrafanaCONline 2021 presentation, Salesforce's team highlighted their use of Grafana’s dashboards, Prometheus, and plugins to derive real-time service health insights and support low-latency alerting with auto-remediation and auto-scaling capabilities. The company's cloud-native architecture relies on hyperlocal observability tools that bundle Prometheus, Grafana, and Alertmanager, working alongside Argus, their time series monitoring platform, to provide a comprehensive alerting solution. The team shared how automation tooling has enhanced the management of alerting and dashboards, offering features like templating, versioning, and integrations. Salesforce's dashboards facilitate trends, health checks, and performance monitoring, while Grafana’s capabilities such as $variables and Javascript callouts enhance usability. This robust monitoring strategy is crucial for Salesforce's Commerce Cloud, which processes vast amounts of e-commerce metrics and alerts, ensuring customers experience a 99.99% platform availability during critical sales events such as Black Friday. The dashboards allow Salesforce to quickly assess system health and steer investigations to diagnose and resolve any customer issues effectively.

Ozark Integrated Circuits has developed a novel approach to monitor ultraviolet radiation on the International Space Station (ISS) using Grafana, an open-source platform for data visualization. The company installed their XNode UV sensor modules on the ISS in January 2019 as part of Alpha Space Test & Research Alliance's MISSE platform to better understand the effects of UV radiation on space materials. These sensors, made from silicon carbide, are designed to withstand extreme environments and provide real-time data that can be analyzed for insights into material degradation in space. The challenge of managing and visualizing large volumes of data from these sensors was addressed by integrating Grafana, which allows researchers to correlate UV data with the ISS's location and other geospatial information. This advancement not only aids NASA and other agencies in understanding UV impacts but also demonstrates Ozark's capability in deploying IoT solutions in extreme environments, with applications extending to deep earth energy exploration.

In July 2021, Grafana Cloud introduced several new features and updates, enhancing its capabilities in observability and monitoring. Grafana Cloud Traces, built on the GA-released Grafana Tempo, allows users to integrate traces with metrics, logs, and dashboards, offering up to 50GB of trace analysis for free. The release of Grafana 8.0 brings improved data visualizations, live streaming, enhanced security, and real-time streaming capabilities for more responsive dashboard interactions. New integrations with services like Kubernetes, Docker, and JIRA aim to streamline the setup of tailored dashboards and alerting systems. Additionally, updated data source plugins, including Humio and Redis, enable visualization of existing data, while a revamped synthetic monitoring plugin offers a comprehensive view of system checks. Grafana Cloud plans to introduce a richer alert management interface in a future release, further aligning with the user-friendly experience of its platform.

Grafana Cloud users often face challenges in managing the increasing complexity of data generated by software systems, particularly in distinguishing which metrics are important and which are unused. To address this, Grafana has introduced commands for their cortex-tools, enabling users to generate a list of unused metrics ranked by cardinality. These unused metrics, which are stored in Grafana Cloud but not displayed on any dashboards, can be identified using the cortextool analyse commands. Users can then remove these metrics by adjusting their remote_write configurations in Prometheus or the Grafana Agent. The goal is to help users manage costs while maintaining essential observability, with Grafana Cloud facilitating this process by providing tools and insights into metric usage.

Amidst challenges in accessing accurate Covid-19 data in Romania, a team from Babes-Bolyai University, including five economists and a data scientist, developed the Romanian Economic Impact Monitor using Grafana to provide a comprehensive dashboard that tracks the pandemic's impact and correlates it with government actions and economic indicators. This initiative emerged as a response to the inadequacy of government-released information, aiming to inform the public effectively through an easily accessible platform. The decision to use Grafana was driven by its ease of implementation and flexibility, allowing the team to create a detailed landing page with layered data, including key indicators like case numbers and economic figures, presented in an intuitive format. Despite not being software engineers, the team successfully operated the site for over a year, making it one of the country's most visited Covid information sources, illustrating the role of universities in crisis communication and public service.