April 2022 Summaries

Theia Scientific and Volkov Labs have collaborated to develop the Theiascope™ platform, an innovative solution that integrates artificial intelligence (AI) with digital microscopy to streamline scientific image analysis. By utilizing the Grafana platform, Theiascope™ facilitates real-time AI-enhanced microscopy, enabling scientists to conduct complex analyses with a single click, which traditionally required laborious manual processes. This integration allows for customizable user interfaces, supporting various experiments and microscopes through a series of custom panels for image acquisition, model management, and real-time quantitation. The application, which has been implemented in institutions like the University of Michigan and Argonne National Laboratory, leverages Grafana's dashboard capabilities to visualize results and manage AI models efficiently. Despite challenges such as time synchronization in restricted environments, the platform promises to enhance the efficiency and accuracy of scientific research by offering an edge computing solution that is both adaptable and scalable.

Deployment-time testing using Grafana k6 and Flagger addresses the challenges of testing microservices in Kubernetes by enabling performance verification before production traffic hits the services. Grafana k6 serves as a versatile performance testing tool capable of running various test cases, while Flagger facilitates the analysis of Kubernetes deployments, supporting canary or blue/green releases and traffic mirroring. The integration of these tools at Grafana Labs involves a Flagger webhook handler that executes k6 scripts, providing a pre-rollout check for each deployment. This setup allows for automated testing, documentation of API endpoints, and observation of results through Grafana k6 Cloud, enhancing deployment reliability and offering valuable insights to developers. The integration is actively being improved, with future enhancements planned based on internal pilot projects and community feedback.

Continuous profiling is an advanced technique for monitoring application performance by analyzing resource utilization over time, allowing developers to identify and rectify performance issues effectively. Originating from Google's 2010 paper on wide profiling for data centers, continuous profiling has gained significant traction in the industry, with many companies investing in this form of observability to enhance system performance. Tools like Pyroscope, an open-source solution, support continuous profiling by providing multi-language support, ad-hoc profiling, and profile sharing, making it accessible and efficient for organizations of all sizes to reduce infrastructure costs and improve scalability. As part of the observability stack, continuous profiling is not only beneficial for frontend developers but is also becoming a cultural norm among engineering teams, as evidenced by industry leaders and events like FOSDEM. Pyroscope offers an intuitive UI and a custom storage engine that can handle various profiling data, enabling businesses to build high-performance applications that enhance customer satisfaction and operational efficiency.

Grafana Labs has introduced a new integration for monitoring Confluent Cloud resources within Grafana Cloud, aiming to enhance observability by visualizing data streaming platform metrics. This integration allows users to connect their Confluent Cloud account to Grafana Cloud with ease, enabling the import of metrics from Kafka clusters, connectors, ksqlDB applications, and schema registries, without the need for agents or custom exporters. The integration is compatible with Prometheus metrics, ensuring long-term data retention and the ability to query using PromQL. Users can utilize a pre-built Grafana dashboard to quickly visualize and filter key metrics such as read/write rates, partition numbers, and connector activity. Setting up involves creating a Confluent Cloud API key with a MetricsViewer role, selecting resources to monitor, and configuring jobs within Grafana Cloud, after which metrics are automatically pushed to Grafana for analysis. This development is part of Grafana Labs' broader effort to expand its platform of integrations, facilitating easier connectivity and monitoring of external systems, and it provides a streamlined way to track Confluent Cloud performance.

Setting up a Spring Boot application with Grafana Cloud, Prometheus, and OpenTelemetry allows developers to effectively monitor and observe application performance by collecting and correlating logs, metrics, and traces. The process involves using OpenTelemetry to instrument a simple Spring Boot application, Hello Observability, with the Grafana Agent facilitating the shipment of this telemetry data to Grafana Cloud. Prometheus is employed to gather performance metrics, which are then correlated with traces using Grafana's exemplars. The Grafana Cloud platform offers comprehensive observability solutions by integrating metrics, traces, and logs without the need for extensive infrastructure management, thereby enabling developers to concentrate on data analysis and business insights. The tutorial also highlights the configuration of Grafana Agent and the use of auto-instrumentation with the Java Agent, which simplifies the process by allowing telemetry data collection without code modifications. With Grafana Cloud's unified observability platform, developers can seamlessly correlate telemetry data, offering actionable insights and facilitating troubleshooting through features like exemplars and the ability to extract metrics from logs.

Grafana Mimir, touted as the most scalable and performant open-source time series database, offers a straightforward migration process from Cortex, as demonstrated in a video tutorial that shows how to transition in under four minutes using Jsonnet to generate the necessary configurations. Initially, Cortex handles the dashboards, but post-migration, Grafana Mimir becomes the backend, showcasing features like query sharding and a two-stage compactor. The tutorial series, which includes "Introducing Grafana Mimir" and "Getting started with Grafana Mimir," supports users in this transition, with further resources available through webinars, documentation, and community discussions. Grafana Cloud facilitates access to these capabilities with a free tier and various plans to accommodate different needs.

Grafana Cloud has enhanced its Kubernetes integration, adding features such as support for Kubernetes events and the ability to collect and ship container logs to Grafana Cloud Logs. This update aims to provide deeper insights into Kubernetes clusters by allowing users to monitor and alert on core metrics using the Grafana Agent. The integration now includes prebuilt dashboards and alerts, alongside reduced metric cardinality to make monitoring more cost-effective. Additionally, the integration supports deploying the Agent as a StatefulSet, preserving its write-ahead log and event history, and features updates from the upstream Kubernetes mixin project. Users can also correlate metrics and logs more effectively through Grafana's Explore view, and future developments promise even more user-friendly navigation and enhanced monitoring capabilities. The integration is available for installation in Grafana Cloud, with documentation provided for updating existing setups.

Blinkit, an eCommerce company offering rapid grocery delivery in India, transitioned from a self-managed Elastic Stack to Grafana Cloud to overcome challenges in its logging processes that were hampering operational efficiency amid rapid growth. The decision to adopt Grafana Loki was driven by its seamless integration with existing open-source tools like Grafana and Prometheus, allowing Blinkit to efficiently handle extensive log data and enhance their monitoring capabilities. This shift enabled the engineering team to focus more on product development rather than maintenance issues. Leveraging the supportive Grafana open-source community, Blinkit also contributed by developing an open-source tool called Legend to facilitate Grafana dashboard creation. Looking forward, Blinkit plans to expand its use of the Grafana Stack with distributed tracing and additional monitoring solutions, aiming to maintain its leadership in the instant delivery market in India.

Grafana 8.5 introduces a range of enhancements focused on improving usability, data visualization, and security, featuring updated panels, new role-based access control (RBAC) options, and simplified reporting features. Notable updates include support for Atom feeds in the News panel, scrollable bar gauge panels, and a new transformation for structuring data in a matrix format. The version also debuts an easy "Explore to dashboard" query workflow, and an expanded navigation menu to streamline user experience. Enhanced tracing functionality now allows linking traces to Splunk logs, while alerting capabilities have been improved with grouped alert rules and fine-grained RBAC, which extends to dashboards, folders, and annotations. The update also introduces better software update notifications and a revamped reporting UI, making report creation more intuitive. Grafana 8.5 aims to provide a more powerful, user-friendly experience for both free and enterprise users, while encouraging community engagement through events such as GrafanaCONline 2022.

GrafanaCONline 2022 promises a robust lineup with over 30 sessions, showcasing innovative uses of Grafana, including the launch of Grafana 9.0. The event, running from June 13 to June 17, will feature projects spanning diverse industries such as monitoring Earth Station Operations for Dish Network, optimizing food waste reduction processes at Apeel Sciences, and analyzing telemetry data for Formula 1 racing using Grafana Cloud and Azure Data Explorer. It will also highlight Grafana's role in supporting IoT applications and the implementation of a dynamic data health scorecard system at Torqata. Attendees can expect deep dives into new features of Grafana 9.0, UX best practices for dashboard creation, and demonstrations from Grafana Labs' company-wide hackathons, including creative projects like real-time drone tracking and running the game Doom on Grafana. Participants are encouraged to register for the free event to explore these sessions and more.

Grafana Mimir, an open-source time series database, introduces a scalable split-and-merge compactor capable of handling up to 1 billion active series, addressing limitations of existing Prometheus and Cortex compactors. The compactor employs features such as grouping input blocks, sharding series, and merging shards to enhance performance and manageability. These features allow for efficient data ingestion, compaction, and deduplication, ensuring fault tolerance and durability through 3x replication. By splitting series into configurable shards, Grafana Mimir overcomes Prometheus TSDB format limits, facilitating better horizontal scaling and improved query performance. The compactor's capabilities allow Mimir to manage the regular influx of data from ingesters, with its design supporting efficient deduplication and block management. The system's flexibility and compatibility with tools like Prometheus and Thanos, coupled with its integration into Grafana's ecosystem, make it a powerful tool for handling large-scale metrics efficiently.

Grafana Labs has determined that it is not affected by the Spring4Shell vulnerability, which involves a potential remote code execution risk in certain versions of the Spring Cloud Function when using routing functionality. Despite the initial report of the vulnerability on March 29, 2022, and its ability to bypass patches for CVE-2010-1622 due to changes in Java Development Kit versions 9 and later, Grafana Labs has thoroughly reviewed its code base and related components, finding no evidence of impact. The company advises users to stay informed through security announcements and updates on their community site and offers a communication channel for reporting any security vulnerabilities. Additionally, updates and recommendations from CISA and VMware on addressing related vulnerabilities in the Spring Framework and Spring Cloud Function are emphasized.

Service level objectives (SLOs) have emerged as a significant focus in the field of observability, transforming how companies like Grafana Labs measure and manage their performance. SLOs provide a framework for defining the quality of service expected over a given timeframe, helping organizations balance the need for rapid innovation with the necessity of maintaining reliability. The adoption of SLOs has been credited with reducing noise in alert systems, improving resource planning, and facilitating more meaningful discussions about service performance. Tools like Pyrra, co-created by Matthias Loibl, aim to simplify the implementation of SLOs by offering configurations that integrate with existing monitoring systems, such as Prometheus. As organizations adopt SLOs, they often experience cultural shifts that emphasize accountability and performance transparency, making SLOs a valuable asset for both internal operations and customer satisfaction.

Grafana Mimir is introduced as a highly scalable and performant open-source time series database, with a video tutorial guiding users through the setup process. The tutorial involves cloning the Grafana Mimir GitHub repository, navigating to the tutorial directory, and starting the cluster using Docker Compose. The architecture includes three Mimir instances, Nginx for load balancing, pre-installed Grafana dashboards, and MinIO for data storage. The system can withstand a single instance failure due to its replication factor of three, ensuring continuous operation and alert notifications through various channels. The video series also covers advanced features like query sharding, and the project hosts webinars for further insights and future developments.

In 2019, Alexander Mann transitioned from the microchip industry to agriculture technology, founding AgriTech after developing a personal IoT setup to manage his garden remotely. AgriTech, based in Virginia, leverages IoT, AI, sensor data analytics, and Grafana to enhance crop health, maximize yields, reduce costs, and lower CO2 emissions for farmers, notably partnering with Pure Shenandoah, a major industrial hemp fiber grower. The partnership has significantly increased hemp production, from a pilot 15-acre project yielding 8,800 pounds in 2020 to over 750,000 pounds from more than 200 acres in 2021, with plans to expand to over 800 acres. Mann's monitoring system, built initially with a Raspberry Pi and Home Assistant OS, uses Grafana to visualize environmental data, aiding in real-time decision-making and alert systems for optimal plant care. Mann aims to educate on industrial hemp's benefits and the positive impact of technology in agriculture, hoping to foster job creation and community support through AgriTech's initiatives.

Grafana Enterprise Metrics 2.0 (GEM 2.0) introduces new capabilities built on top of Grafana Mimir 2.0, including the experimental feature of cross-tenant alerting and recording rules, which address the complexity of managing large-scale alerting and data aggregation within organizations. This feature allows for the creation of a single rule group that can access and operate on data from multiple tenants, avoiding the redundancy of deploying rules separately for each tenant and reducing the administrative burden. Cross-tenant rules provide a unified view of metrics across different teams, facilitating alerts based on aggregated data, such as monitoring CPU usage or database errors across multiple tenants. GEM 2.0 integrates this functionality with its existing access control system by ensuring that access policies are checked during rule evaluations, maintaining secure and authorized data access. Although creation and modification of these rules currently require using mimirtool or the ruler API, future plans aim to enhance user experience by enabling them through the Grafana Alerting UI. This release aims to simplify tenant management and improve the efficiency of monitoring and alerting processes across complex infrastructures.

Grafana Enterprise 8.4.6 has been released to address a high-severity security vulnerability (CVE-2022-24812) affecting versions 8.1.0-beta1 through 8.4.5, particularly impacting instances with the fine-grained access control beta feature enabled and multiple API Keys with different roles. Discovered during an internal audit, the vulnerability could lead to privilege escalation, prompting an urgent need for users of the affected versions to upgrade or disable the feature. Grafana Cloud instances have already been updated, and major cloud providers such as Amazon and Azure Managed Grafana have been notified and confirmed secure. The timeline of the security issue outlines the swift response from discovery to public release, and Grafana Labs provides a secure channel for reporting vulnerabilities, requesting non-disclosure until fixes are announced. Security updates are communicated via the Grafana community site and blog, offering email and RSS subscriptions for ongoing notifications.

Grafana Loki 2.5 introduces significant performance enhancements, expanded log ingestion options, and new operational features, including nearly 500 pull requests since the previous version. Key improvements include optimized regular expression handling, faster binary operations, and a new storage schema to overcome S3 rate limits, allowing more efficient queries. New log ingestion methods have been added, such as service discovery and API tailing from the Docker Daemon, direct log fetching from Cloudflare, and support for receiving Graylog Extended Log Format messages. The version also features an improved Simple Scalable Deployment (SSD) mode, facilitating easier management on both Kubernetes and standalone environments, with new Helm charts available for Grafana Enterprise Logs. Additionally, the update introduces anonymous usage reporting to guide future development, while maintaining user privacy, with options to participate or disable this feature. These changes aim to enhance the overall user experience and operational ease, ensuring Loki remains a leading solution for object storage logging.

Grafana Labs introduced Grafana Mimir, an open-source time series database (TSDB) designed to handle up to 1 billion active series, effectively addressing the increasing demand for scalable metrics monitoring solutions. Extensive load testing in 2021 led to significant optimizations in Grafana Mimir's scalability and performance, resolving challenges such as CPU utilization, compaction inefficiencies, and query execution times. Key improvements included reducing CPU usage for hash rings, enhancing the compaction process with a split-and-merge algorithm, and implementing query sharding to handle high-cardinality queries efficiently. The deployment involved a robust architecture with thousands of replicas and terabytes of memory and CPU resources, ensuring efficient data ingestion and querying. Grafana Mimir's enhancements, including asynchronous chunk writing and multi-zone rollouts, have significantly reduced latency and improved deployment consistency, making it a powerful tool for organizations dealing with extensive time series data.

Grafana Labs has secured a $240 million Series D funding round led by GIC, with participation from new investor J.P. Morgan, marking a significant milestone for the company. This investment, just seven months after their Series C, is seen as a testament to Grafana Labs' product strategy, execution, and strong customer base. GIC, having initially invested in the company's Series B round, aims to support Grafana's global growth and deepen its presence in the APAC region. J.P. Morgan, whose engineers have been actively involved in Grafana's open-source community since 2018, sees both short-term and long-term potential in the partnership. The funding will be used to enhance Grafana's product offerings, including open-source projects like Grafana Loki, Grafana Tempo, and Grafana Mimir, which emphasize the company's commitment to innovation and community engagement. Grafana's "big tent" approach aims to empower users with technology to control their observability strategies, as evidenced by diverse applications ranging from satellite launches to beekeeping.

WebSockets offer a seamless way to visualize real-time IoT data in Grafana, particularly when used with Golioth's IoT development platform. The platform facilitates secure connections to numerous devices, allowing data collection and integration with various systems. Golioth has developed an open-source Grafana plugin that connects to any WebSocket data source using JSON format, enabling automatic updates of Grafana panels as new data arrives without polling. This feature supports both continuous data streams and infrequent alert data, providing real-time monitoring capabilities. The integration process involves adding the WebSocket plugin to Grafana, configuring the WebSocket URL and API key, and setting up a panel to display the incoming data. By using this setup, users can achieve immediate feedback on the health of their devices, making it an essential tool for managing remote IoT hardware fleets. The Golioth platform, which offers a free Dev Tier, allows users to quickly bring devices online, generating data and demonstrating the plugin's capabilities.

"Grafana's Big Tent" is a podcast initiated by Grafana Labs to expand discussions around observability beyond internal conversations, featuring hosts Matt Toback, Mat Ryer, and Tom Wilkie. The podcast aims to foster inclusive dialogues about technology, community, and tools, engaging both community members and those from competing platforms. Episodes, released bi-weekly, explore modern observability and include candid discussions about successes and failures in using various tools, emphasizing learning from mistakes. Grafana's approach is to integrate diverse ideas and user practices rather than mandating changes, reflecting their philosophy of inclusivity and adaptability. The podcast encourages listener participation and feedback, further highlighting its commitment to community engagement and innovation.

Grafana Labs is launching a new webinar series in the Asia Pacific region to help users, irrespective of their experience level, get started with the Grafana Stack for observability, focusing on metrics, logs, and traces. Throughout April, the series will cover various aspects such as setting up Grafana with Prometheus and Grafana Loki for monitoring web services, exploring Grafana's scalable solutions for integrating metrics from multiple systems, utilizing Grafana Loki for efficient log management, and using Grafana Tempo for high-volume distributed tracing. Each session will include live demonstrations by Grafana Labs technical experts, showcasing the latest features and enhancements in Grafana tools, as well as opportunities for participants to ask questions. The initiative reflects Grafana Labs' commitment to providing open-source solutions that integrate seamlessly with users' existing tools and infrastructure.