Bugs, Open Source Software and the Cloud
By Tim Hall / Mar 16, 2017 / InfluxDB Cloud, Developer
Yesterday, we announced that InfluxDB Cloud, our managed database as a service offering, has been upgraded to use the latest InfluxDB Enterprise version 1.2.2 (also released yesterday). As prospects register for trials and new customers sign-up, all new instances of InfluxDB Cloud leverage the latest innovations and capabilities from across the open source community, and we continue to upgrade our existing InfluxDB Cloud customers to this latest release. More than half have already been upgraded and we are continuing to roll out this update to the rest of our customer base. We are very excited about this latest release and you might ask yourself…how can you possibly be excited about what appears to be a maintenance release?
One of the promises of open source software is that quality tends to be higher than proprietary software. An active and vibrant open source community means that the software being constructed is used extensively, and therefore tested, in a wider variety of ways than a single software vendor might imagine – exercising more of the possible code paths that exist in these ever evolving and complex technologies. I’d like to take a moment to thank everyone within the open source community who contributes, collaborates, and continues to help build the TICK stack. We are truly appreciative of every issue logged and pull request submitted. These things help us build and deliver a product that continues to delight developers everywhere.
While InfluxDB Enterprise is based on the open source TICK stack, it also includes closed-source capabilities such as high-availability (HA) for InfluxDB and more. Our customers continue to benefit from the testing and exploration done on the underlying components by the open source community, but the question for InfluxData is how best to battle test InfluxDB Enterprise? That is where InfluxDB Cloud comes in.
As TICK stack components are developed and released within the open source community, that triggers the beginning of our activities to deploy and operationalize those bits as part of an upcoming InfluxDB Enterprise release. There is a lot of work going on behind the scenes at InfluxData. In the same way that the open source community helps with the shake-out of new features and functions, we are able to leverage the wide variety of customer deployments, setups, and scenarios running within the InfluxDB Cloud to help us further harden the release, focus on the unique capabilities of the closed source bits, and ultimately deliver a higher quality release – for both our on-premise and cloud customers. It is inevitable that there will be software defects uncovered and we will continue to work diligently to address these quickly. But, we are focused on continuing to drive the quality of our software higher. There are some really innovative ways that we can take advantage of the variety of deployments and operational situations while running InfluxDB Enterprise within the InfluxDB Cloud – and we plan to expand our efforts in this regard as we deliver new releases.
As we run, manage, and monitor InfluxDB Cloud on behalf of our customers, we are working hard to continue to improve the stability, reliability, and the overall experience of the service. The “digital exhaust” generated as customers use InfluxDB Cloud, such as operational statistics, logs, and other monitoring data, is invaluable in helping our engineering team understand where to focus our efforts and improve InfluxDB Enterprise. The bulk of the instances in InfluxDB Cloud leverage the HA setup, and as we perform maintenance and upgrades of the service, we leverage these capabilities in real-world scenarios regularly – ensuring uninterrupted access to the service. Further, when dealing with distributed systems, having long burn-in cycles allows us to monitor, detect, and address things like memory leaks or excessive file handles being used, and more. The InfluxDB Cloud also provides us with the opportunity to review and explore optimizations in configuration settings along with nuances in different classes of compute and storage hardware. The benefit to our customers is that we can help de-risk new features, reduce regression-related issues, simplify the upgrade process, understand and address operational challenges before they manifest themselves within customer environments, and make the process of adopting new releases as frictionless as possible.
So, we are thrilled to deliver this latest version of InfluxDB Enterprise. What may seem like a minor release has gone through extensive testing by the community, by our team, within the cloud, and is ready for customers to deploy. Again, thank you to everyone involved in driving the quality and reliability of this release!