November 2016 Summaries

The TICK Stack, which includes Telegraf, InfluxDB, Kapacitor, and Chronograf, is a comprehensive monitoring solution for Kubernetes clusters. This stack was created to provide an easy-to-use and scalable way to monitor and manage data in real-time. The TICK Stack uses Helm Charts to deploy the components, making it easy to set up on different Kubernetes configurations. Telegraf collects metrics from the cluster, which are then forwarded to InfluxDB, a time-series database that provides a scalable and reliable solution for storing and managing data. Kapacitor is used to process and analyze data in real-time, while Chronograf provides an intuitive web interface for visualizing and managing data. The TICK Stack can be deployed on-premises or in the cloud, offering flexibility and scalability for organizations of all sizes.

InfluxDB has released InfluxDB 1.1, which offers up to a 60% performance increase compared to its predecessor. This new version also introduces new query functionality and improved support for Kubernetes monitoring and autoscaling with Telegraf and Kapacitor. The company is also adding multi-tenant support for Grafana on InfluxDB Cloud, allowing customers to configure access to different groups and users in their organization. Additionally, InfluxDB has been marked as the winner of a time-series data and metrics benchmark against OpenTSDB, showcasing its capabilities in this area. Furthermore, the company is hiring and offering various training sessions, including webinars, virtual trainings, and hands-on workshops, to help customers get the most out of their InfluxDB setup.

In the latest release of InfluxDB, version 1.1, several new features and syntax developments were introduced, including syntax consolidation for SHOW queries, which now support the ON <database_name> clause and are optional. This change ensures consistency in query syntax across different queries. Additionally, two new functions were added: SAMPLE(), which returns a random sample of N points from specified data, and CUMULATIVE_SUM(), which calculates the cumulative sum of consecutive field values. These updates aim to improve performance, stability, and user experience. The release also includes other improvements and features, such as downloads for the TICK-stack, cloud deployment options, and a call to action for users to share their InfluxDB success stories.

InfluxDB has recently released version 1.1, which offers up to a 60% performance increase and new query functionality. The company also announced the new Chronograf UI, a complete open source monitoring solution that is part of the TICK Stack. Additionally, InfluxDB Cloud now supports multi-tenant access for Grafana, allowing users to configure access to different groups and users in their organization to view information with their own dashboards. Made.com uses InfluxDB as the heart of their monitoring system to store metrics gathered from their Collectd and Riemann stack. The company is also hiring and offering free virtual training sessions on performance tuning and schema design, as well as an intro to Telegraf and the plugin ecosystem.

In the latest release of InfluxDB, version 1.1, several new features have been introduced to improve performance, stability, and query capabilities. The database now supports mixed duration units in retention policies, continuous queries, and time conditions, allowing for more flexibility when specifying durations. Additionally, a configuration setting called `max-values-per-tag` has been added to limit series cardinality, preventing excessive tag values from being written. Furthermore, InfluxQL support for linear interpolation in GROUP BY time() queries has been enhanced, enabling users to specify the interpolation method for filling gaps in data without any data points. These updates aim to make working with InfluxDB more efficient and user-friendly.

The new Chronograf, part of the TICK stack, offers an open source monitoring solution that provides developers with flexibility to customize their visualizations and monitoring algorithms. The goal is to give users the power to deploy a cost-effective solution while giving them control over their monitoring. This release marks a significant step towards achieving this vision, providing a user interface for the entire stack as a complete open source monitoring solution. With features like Kapacitor integration, alerting rules, and community contribution through plugins, Chronograf aims to make it easier for users to create complex monitoring systems without needing to write TICK scripts. The release is considered an alpha version, with plans for frequent updates based on user feedback.

InfluxDB 1.1 has been released with significant performance improvements, including up to a 60% increase in query performance for first() and last() queries without GROUP BY, as well as substantial increases in write throughput and improved stability for sparse high cardinality data sets. The release also introduces new query capabilities such as linear interpolation, cumulative sums, support for regular expressions on field keys, and the ON clause for all SHOW commands. Additionally, the admin UI is deprecated and disabled by default, while upgrading from 1.0 to 1.1 requires no data migration but may require configuration changes prior to upgrade. The release highlights InfluxDB's commitment to iterating quickly based on community feedback.

InfluxDB has been actively releasing new content, including webinars and training sessions, to help users improve their skills and knowledge of the platform. The company has also made several announcements, such as adding multi-tenant support for Grafana on InfluxDB Cloud, which allows customers to configure access to different groups and users in their organization. This feature is particularly useful for companies that collect powerful data about their applications and services. One notable example is Citrix Systems, which uses InfluxDB to track incoming bugs and actions from JIRA and Github, making it a valuable tool for developer standups. Additionally, the company is hiring new talent and has announced several events, including free virtual training sessions and hands-on workshops, to help users learn more about the platform.

The Kubernetes platform has added new metrics capabilities, made possible by the contributions of its open-source community. These new features allow for more detailed insights into a Kubernetes cluster's performance, such as request per second and queue depth. The Telegraf plugin developed by Jonathan Chauncey extends the type of metrics gathered to include these additional key performance indicators. As a token of appreciation, the project is sending out t-shirts to contributors and inviting them to share photos with the shirts on social media. The community's contributions are essential to the platform's growth and success, making this recognition a well-deserved shoutout.

In this post, we recap the most interesting timestamp and query time range issues, workarounds, and how-tos from GitHub, IRC, and the InfluxDB Google Group in the last week or so. The maximum timestamp that InfluxDB can handle is 2262-04-11T23:47:16.854775806Z, while the minimum timestamp is -9223372036854775806 or 1677-09-21T00:12:43.145224194Z. When a query doesn’t specify a time range in the WHERE clause, InfluxDB covers a default time range that varies depending on the version of InfluxDB being used. For example, in versions 1.0 and under, the default time range is between 1677-09-21 00:12:43.145224194 UTC and now(), while in versions 1.1+ the default time range is between 1677-09-21 00:12:43.145224194 UTC and 2262-04-11T23:47:16.854775806Z UTC. Additionally, users can configure returned timestamps by using the epoch query string parameter in the HTTP API or specifying alternative timestamp formats with the precision command in the CLI.

InfluxDB has achieved writes greater than 1 million values per second in a minimal configuration of 4 m4.4xlarge instances, with the ability to sustain this performance through batched HTTP protocol usage and without requiring triple replication of data, resulting in disk space savings. The company tested various configurations on AWS M4 instances, including a larger cluster that reached 1.5 million writes per second, demonstrating InfluxDB's potential for high-performance distributed databases. This achievement is notable as it surpasses common benchmark goals in the industry and sets InfluxDB apart from other distributed databases.

InfluxData has released Telegraf and Kapacitor versions 1.1, which improve Kubernetes cluster monitoring and autoscaling. The new releases provide a more efficient use of InfluxDB's storage backend and better integration with the Kubernetes community. Telegraf now supports gathering a wide range of Kubernetes metrics, including those provided by Heapster, allowing for a complete view of the data. Kapacitor 1.1 enables easy scaling of Kubernetes clusters using any data stored in InfluxDB, providing more flexibility than the traditional horizontal pod autoscaler. The TICK-stack is now available for download, and users can deploy it on cloud or server environments with free trials. Additionally, the community is encouraged to share their success stories and provide feedback on future developments.

InfluxDB has been actively engaging with its community, releasing new features such as multi-tenant support for Grafana on InfluxDB Cloud, which allows configurable access to different groups and users in their organization to view information with their own dashboards. This feature was made possible by customer feedback, indicating a strong demand for this functionality. Additionally, various success stories have been shared from customers who have successfully implemented InfluxDB in production, including ReproConnect, Happiest Minds, and Mas Ingenieros, highlighting its capabilities in handling large amounts of data, providing customizable dashboards, and enabling real-time monitoring and alerts. InfluxDB also announced new training opportunities, including virtual workshops and webinars, to help users get the most out of their platform. Furthermore, the company is inviting companies that use InfluxDB in production to claim a hoodie and be featured on its website and social media channels.

We're going to GrafanaCon 2016` GrafanaCon 2016 will be held at the Intrepid Sea, Air & Space Museum on the Hudson River in NYC, where attendees can meet Paul Dix, the CTO of InfluxData. The event aims to hear from users about their experiences with InfluxDB and Grafana, with a raffle for a free pass to the conference. To inspire users, recent examples of solutions using InfluxDB and Grafana are being presented, including a currency dashboard, a simple monitoring stack setup using Docker, and a platform for monitoring anomalies in experimentation platforms. These examples showcase the versatility and potential applications of InfluxDB and Grafana.

The InfluxDB Tech Tips article discusses various ways to interact with InfluxDB, including writing data from a file and exporting data in line protocol format. The article also addresses common issues such as setting an infinite shard group duration, which is not currently supported by InfluxDB. Instead, users can specify a very long shard group duration using the 1000w value. Additionally, the article provides information on available tools and resources for InfluxDB users, including the influx_inspect tool and support options.