May 2016 Summaries

In this post, we recap the week's most interesting InfluxDB Kapacitor and TICK-stack related issues, workarounds, how-tos and Q&A from GitHub, IRC and the InfluxDB Google Group that you might have missed. When using an INTO query to move data to a new measurement, it's essential to use GROUP BY * to preserve tags as tags instead of converting them to fields. Streaming or joining stream data with shifted stream data in Kapacitor can be problematic for large time ranges, and it's recommended to use batch tasks instead. InfluxDB queries also require double-quoting identifiers that are also InfluxQL keywords.

InfluxDB has announced early access to its Enterprise version, which enables clustering on infrastructure, allowing for rebalancing of nodes and deployment management through a slick UI. The company also released free virtual training sessions, including one on migrating from InfluxDB 0.8 to 0.13, and several tech tips articles. Various companies, such as SolarCity, DNSFilter, and Commonwealth Financial Network, have successfully implemented InfluxDB in their production environments for real-time data collection and analysis, with benefits including improved scalability, monitoring, and decision-making capabilities. InfluxDB is also being featured on its website to promote user success stories, and the company is hosting free meetups and training sessions in the coming weeks.

InfluxDB is a time-series database that provides a way to query series cardinality, which refers to the number of unique values in each series. The series cardinality can be found using a specific SQL query. Additionally, Kapacitor TICKscript filters can be used to define multiple levels of filtering, but Kapacitor assumes that each level is a subset of the previous one, resulting in unexpected behavior if not designed correctly. To resolve this issue, filter criteria should be adjusted so that the WARN level includes values for the CRITICAL level. Furthermore, InfluxDB version 0.13 introduces a feature to log slow queries after a certain threshold, which can be configured using the `log-queries-after` setting in the configuration file.

InfluxDB has recently released early access to its Enterprise version, which allows users to run clustering on their own infrastructure with rebalancing and UI management capabilities. This new feature is designed for organizations that require more control over their data storage and management. Several companies have successfully adopted InfluxDB, including Endticket, RedBlink, and Hoiio, who use it for various applications such as ticket sales platforms, music streaming apps, and monitoring systems. These users praise InfluxDB's scalability, ease of configuration, and performance in handling large amounts of data. The company is also offering a free virtual training session on comparing InfluxDB with Elasticsearch for time-series data management, and providing official Docker images for its components. Additionally, InfluxDB has been featured in various benchmarks, including one that shows it outperforms Elasticsearch in time-series data and metrics management.

In this article, we recap the week's most interesting InfluxDB and TICK-stack related issues, workarounds for customizing GROUP by time() windows and series cardinality, and Q&A from GitHub, IRC, and the InfluxDB Google Group. Regular expressions in the WHERE clause can be used with proper operators, such as =~ for positive matches or !~ for negative matches. Additionally, retention policies do not affect series cardinality, as the series index only applies to unique measurement and tagset combinations. Furthermore, time windows returned by a GROUP BY time() query can be customized using an offset interval starting from InfluxDB version 0.13. These customizations allow users to fine-tune their queries for better performance. For more information on these topics and others, the article provides links to frequently encountered issues, virtual training seminars, and other resources.

The team at InfluxData is excited to announce the immediate availability of InfluxDB v0.13, the final major milestone before the launch of version 1.0, delivering robust features and stability to the single server experience. This release includes 27 new features and 32 bug fixes, as well as upgrades to Go v1.6.2 for improved performance, support for DELETE FROM queries, and time zone support. Additionally, InfluxDB Enterprise is now available in early access, offering clustering capabilities and a user-friendly interface for deployment, management, and monitoring on infrastructure. Furthermore, the company has launched new technical papers, virtual training sessions, and free consultations with solutions architects to help developers and architects get started with InfluxDB v0.13 and explore its features and capabilities.

Kapacitor is a powerful data processing engine in the TICK stack, offering flexibility and customization options for performing complex tasks beyond what Continuous Queries (CQs) can achieve. Kapacitor can process both stream and batch data from InfluxDB, allowing users to plug in custom logic or user-defined functions to transform and analyze their data. The choice between using Kapacitor instead of CQs depends on the specific use case, such as isolating workload, performing more complex data transformations, downsampling for retention policies, or having a small number of CQs. Stream tasks are suitable for scenarios with limited RAM and short time periods, while batch tasks are better suited for larger datasets and longer retention periods. Kapacitor offers the same functionality as CQs but provides more flexibility and power, making it an attractive alternative for complex data processing needs.

InfluxDB has been featured in various projects such as ioWaves, InduWaves, Zalando, and Madberry, showcasing its ability to store time-series data efficiently and provide a clear query language. InfluxDB's simplicity of installation, schema discovery features, and speed were key factors for Zalando's adoption. Additionally, Madberry leverages Telegraf, InfluxDB, and Kapacitor to monitor their systems and ensure 24/7 uptime. The company also offers free virtual training on Telegraf and plugin ecosystem, as well as a hoodie giveaway to companies using InfluxDB in production.

In this post, we recap the week's most interesting InfluxDB 0.13 and TICK-stack related issues, workarounds, how-tos and Q&A from GitHub, IRC and the InfluxDB Google Group that you might have missed. Series cardinality for a single measurement is actually the number of tag sets that exist, not just the product of the number of tag values per tag. You can use the INTO query to move data from one database to another in InfluxDB. A new DELETE series functionality has been added in InfluxDB 0.13, allowing users to delete data based on their timestamp using the WHERE clause. The new feature supports time intervals and does not drop series from the index like DROP SERIES does.

InfluxDB has been featured in various articles, blogs, videos, and tutorials over the past week, including announcements about new features and services such as free virtual training sessions on performance tuning and schema design, fully managed InfluxDB clusters with Grafana on AWS, and time-series monitoring with Telegraf. Several companies, including Onefootball, CrossEngage, and SpareFoot, have successfully implemented InfluxDB to monitor their infrastructure and application metrics, citing its ease of use and powerful features such as high write throughput and easy scaling abilities. Additionally, InfluxDB has been featured in various events and meetups, including virtual training sessions and a meetup in San Francisco, with opportunities for organizers to get support from the company.