InfluxData has bolstered the query performance of its time-series database running on the Amazon Wed Services (AWS) cloud service with the addition of a read replica capability that automatically makes additional compute capacity available when needed.
Integrated into the AWS Management Console, Amazon Timestream for InfluxDB Read Replicas is a commercial add-on to an open source time-series database that AWS made available last year.
InfluxData CEO Evan Kaplan said the Amazon Timestream for InfluxDB Read Replicas offering ensures access to critical time series data is always available across multiple AWS zones without requiring an internal IT team to manually provision additional cluster capacity
In addition to providing an automated failover capability, Amazon Timestream for InfluxDB Read Replicas makes it possible to automatically load balance high-volume queries using read replicas in a way that ensures more compute capacity is available for writes to the InfluxData database, noted Kaplan.
As a provider of an open source database, InfluxData has a unique relationship with AWS, said Kaplan. InfluxData allows AWS to host its open source database under a permissive license while it in turns focuses its efforts on making available additional commercial services. That approach enables InfluxData to generate enough revenue to continue to fund the development of an open source database, he added.
That’s critical because other providers of open source software have decided that rather than allow a cloud service provider to generate revenue by hosting open source software, they opted to change the licensing terms under which they make their software available. Those changes, however, often result in the open source community that developed around that software banding together to create a fork of the project that winds up pulling maintainers and contributors away from the original project.
In some cases, that shift has resulted in the original vendor that sponsored the open source project being acquired when they could no longer generate enough revenue from support services to continue to grow.
Interest in time-series databases, meanwhile, only continues to grow as the volume of telemetry data that IT teams need to collect and analyze grows alongside the number of applications that need to run in near real-time. A time-series database makes it possible to natively access that data versus storing it in a data lake that requires IT organization to then convert it into a table format using a set of extract, transform and load (ETL) tools, noted Kaplan.
In addition, the number of artificial intelligence (AI) applications that need to access time-series data in near real time has sharply increased, he added.
It’s not clear how many organizations are adopting time-series databases, but it’s clear there is a greater need. Most organizations today routinely deploy multiple types of databases in support of a range of applications that all need to be managed by an IT team. The challenge, as always, is finding the expertise required to manage each type of database added to distributed computing environments that only continue to become increasingly more complex with each passing day.
