Three Ways InfluxDB Enables You to Use Time Series Data Across Your Entire Engineering Team
Session date: Apr 21, 2021 08:00am (Pacific Time)
The more your team can collaborate around data, the more useful that data is. This is especially true for time series data that is increasingly the heartbeat of your business. When your entire team can utilize time series data, they know the pulse of your devices, your equipment, your customers, and your software - and can act accordingly.
In this webinar, Russ Savage - product manager at InfluxData - will show you three new ways for your team to collaborate around time series data.
- First, InfluxDB Notebooks let you create and share computational narratives that combine live code, visualizations, and explanatory notes, which can output to your InfluxDB dashboards, tasks, and buckets. You can use Notebooks to better document your downsampling, data processing, incident investigations, postmortems and runbooks.
- Next, InfluxDB Annotations let you explain the why behind time series data trends. Annotations can be used to communicate how time series data is impacted by changes to software deployments (like configurations, upgrades, or outages), user behavior (Cyber Monday, deadlines), business activities (ad campaign, sales incentives), or external events (natural disasters, weather). With team members sharing contextual clues, you'll more quickly determine root cause and restore services faster.
- Finally, learn how to apply GitOps practices to manage InfluxDB configurations, dashboards, tasks, and alerts, as well as Telegraf configurations, ensuring better collaboration workflows between developers, SREs, and every stakeholder involved in time series collection, enrichment and analysis.
Sign up now so you can help your entire team leverage the trio of capabilities profiled above and get the most out of the time series data you’re storing in InfluxDB.
Watch the Webinar
Watch the webinar “Three Ways InfluxDB Enables You to Use Time Series Data Across Your Entire Engineering Team” by filling out the form and clicking on the Watch Webinar button on the right. This will open the recording.
[et_pb_toggle _builder_version=”3.17.6” title=”Transcript” title_font_size=”26” border_width_all=”0px” border_width_bottom=”1px” module_class=”transcript-toggle” closed_toggle_background_color=”rgba(255,255,255,0)”]
Here is an unedited transcript of the webinar “Three Ways InfluxDB Enables You to Use Time Series Data Across Your Entire Engineering Team”. This is provided for those who prefer to read than watch the webinar. Please note that the transcript is raw. We apologize for any transcribing errors.
- Caitlin Croft: Customer Marketing Manager, InfluxData
- Russ Savage: Director of Product Management, InfluxData
Caitlin Croft: 00:00:03.862 Hello, everyone, and welcome to today’s webinar. My name is Caitlin Croft, and I’m very excited to be joined by Russ Savage who’s part of our product management team. And he’ll be providing a product update on some of the really cool stuff that we’ve been working on like InfluxDB Notebooks, InfluxDB Annotations, and GitOps. Once again, this session is being recorded and will be made available. Please, feel free to post any questions you may have in the Q&A. All right, Russ, over to you.
Russ Savage: 00:00:42.214 Thank you, Caitlin. Good morning, everybody. Well, it’s good morning to me, it’s a hello to everyone out there. I appreciate you joining this webinar today. We’re going to talk about three ways that InfluxDB enables you to use your time series data across your entire engineering team. So really exciting topic, and I’m really pumped to share what our engineering team has been working on behind the scenes to enable sharing of data and collaboration. I can’t promise that there’s going to be fake rubber masks that I pull off, like the Apple event yesterday, but it’s going to be exciting nonetheless, so I’m looking forward to getting started. My name is Russ Savage, I’m a product manager here at InfluxData, I work directly with the engineering teams to build some of this amazing technology, and so I’m really excited to share it with you. So just real quick, everyone is at an InfluxData webinar, so I assume that you have a pretty good idea of what InfluxDB does. But for those of you that don’t, the short and sweet summary, InfluxDB Cloud is a time series platform for modern application builders. So if you have a time series data problem, if you have an application that relies on data with a timestamp, InfluxDB Cloud exists to make it easier for you to build those applications. We’re an insanely fast database for ingestion and querying, we include a ton of analytics and processing tools built right in to make it really easy to work with your data, and really easy to bring that data into your application, use it to build awesome experiences for your customers. So InfluxDB Cloud is a time series platform for modern application builders.
Russ Savage: 00:02:34.398 And so today, we’re talking about collaboration. Today, we’re talking about sharing and using your time series data across your organization team or your engineering team. And what we found when we’ve talked to customers, when we talked to our users, as if developers today, they rarely build applications alone. So I think when you think about the developer sitting alone in a dark room hacking away a code, that’s not how a lot of developers are building applications today. The size and scale of the applications that they’re developing necessitate teams, necessitate groups of people working together to solve really interesting problems. And developers collaborate a bunch of different ways across a ton of different tools. So I myself, I’m a product manager now. I’m a recovering developer. I have used all of these tools and many more. I’m sure that any developers out there are very familiar with this. You run into a really tricky problem when you’re writing an application, forums, and discussion boards, a very common way that developers collaborate, you’ve got Slack workspaces, you’ve got Discord servers, you’ve got all sorts of different ways where people can share common patterns, common code patterns, help collaborate on solving different problems; obviously, anybody who’s worked in a team or worked at a company, they’ve used code reviews and pull requests in order to provide ways to collaborate on code that’s being checked in. I know that every time I do a PR, I have hundreds of comments, it’s kind of a badge of honor, but really a powerful way for developers to collaborate and communicate on code that’s being checked into a shared application.
Russ Savage: 00:04:25.789 When you go to deploy your application, deploying applications usually requires collaboration with many different teams: infrastructure teams, security teams, different SRE teams, and so there’s collaboration required to bring that application out and actually put it into a production environment. And then once your application is out there, and probably long before your application is out there, operational monitoring is usually involved. You’ve got teams of people looking at dashboard, looking at metrics about your application, and all of the services that it relies on to make sure that it’s operating at peak performance and there aren’t any outages, right? And so this is not all the ways that people collaborate. It’s just a handful of different ways that - thinking about the development lifecycle of building an application where teams or groups of developers might collaborate. And so we sit back and we think, “Okay, well, if InfluxDB Cloud is a time series platform for application developers, and developers rarely build alone, and they’re looking for ways to collaborate, how can InfluxDB Cloud help developers collaborate?” right? So what are the tools that we have available to make collaboration when you’re building applications about time series data easier? And when I say easier, I mean it’s faster, you write less code, you’re able to write less boilerplate code, focus more on the application code or the or the logic that really drives value to what you’re building, and so we think about all the different ways that we could help developers collaborate, and I’m going to share a couple that we’ve been building and one that we’ve had for a while that I think deserves special mention.
Russ Savage: 00:06:18.166 So InfluxDB Cloud platform comes with a really awesome web interface. So if you haven’t already signed up and logged in, everything is for you to try out, you get a free forever account, just use your email to sign up. You get access to a web interface, really powerful, provides a ton of capabilities out of the box, so this is a screenshot here of just a dashboarding capability, which is just a small piece of the overall capabilities of our web interface to the platform, so it includes a ton of tools to help developers build and monitor their applications, and all of it is geared around increasing the speed of development. So just a couple of things that I want to touch on that it includes. So our web interface has a complete dashboarding system, it’s really, really powerful for operational monitoring of the different applications and different services that those applications rely on, so you can log in and your team of developers that are supporting that application can see real time metrics about what’s happening in the InfluxData system but also in the application and the code that you build. Behind the scenes, one of the most powerful capabilities of our cloud platform is the ability to schedule a background processing tasks for the data that’s being ingested, right? And so a lot of developers are leveraging this capability to do things like intelligent roll-ups and downsampling, things like data enrichment or processing or cleaning up of data as it comes in, and so our web interface provides a way to build those background processing tasks completely in your browser, deploy them and schedule them, and then monitor them as well.
Russ Savage: 00:08:12.959 And the last piece is our web UI really provides an ability to install what we call InfluxDB templates. They’re pre-configured monitoring solutions for common technology stacks. And so think about if your application is - if you’re building an application and it relies on a Kafka queue for processing and incoming data, right, there’s an InfluxDB template out there for monitoring Kafka. It is built by the community, it’s improved by the community, and so it has a collective knowledge and understanding of the important metrics that are used to monitor Kafka, and you can install that into your InfluxDB Cloud account and be monitoring Kafka in a matter of minutes, so really, really powerful tool, and I’ll talk a little bit more about how that relates to kind of the GitOps workflows that we’re going to discuss a little bit later. All right. So the InfluxDB Cloud web interface, it’s truly a powerful tool for teams and developers to build their time series application, and so we want to think about the different ways that we can support collaboration through that tool. So another big word slide, I may have gone overboard on these giant word slides, but hey, they’re included in our marketing template, so why not use them? So how can we support collaboration while you’re actually monitoring your applications? A lot of people are leveraging on our web interface for dashboarding capability, so what are the different ways that we can help support that?
Russ Savage: 00:09:56.264 Now, obviously, when you build a dashboard, you yourself can look at it, but then anyone in your organization, anyone in your team can log in and view that information, view the graphs and the data that you’ve used, there’s a ton of tools in our dashboarding platform around adding notes and comments to individual dashboards, but just recently, we’ve added the ability to add Annotations directly to the data being displayed on the dashboard. And so in this case, I’m looking at some data from a national buoy database, that’s data being streamed in from all of those different buoys that are blinking out in the oceans around the world, that data is available for free, so I’m pumping that into my InfluxDB instance. And so here, I was looking through some data, I noticed there’s a big spike in the significant wave height right off the coast of the northeastern United States. So it’s an anomaly, it seems that it’s usually at about a meter, looks like it’s spiked to nearly seven meters, that’s quite a difference, 7X increase in wave height might be something to look into and investigate, and so I’m able to quickly enable Annotations, add an annotation for my teammate, Greg, to have him take a look.
Russ Savage: 00:11:26.385 So you can use this to add notes to the time series data that you’re leveraging on the platform; it’s incredibly handy for highlighting deployments, highlighting outages, highlighting service downtime, highlighting different anomalies that you’re investigating. If you’re a group of people or you’re a team of people, sometimes, especially now, you’re not always in the same place, but a lot of times, you’re looking at the same dashboards, so you can quickly see something, let your teammates know that you’re actually actively investigating it, and so anytime that they come and look at the dashboard, you don’t spend duplicate time investigating the same spike. Super easy to use. All you have to do, there’s a button at the top of every dashboard. Click the Annotations tab. Click Enable Annotations, and start annotating your data. Really powerful way to add context to your team members, obviously, the basics, add or delete as needed. The Annotations are stored separate from your time series data, so they never expire. Your Annotations will stick around until you delete them, which is fine. There’s no charge for adding Annotations, so add as many as you want in your cloud account, it’s super awesome. They’re coming soon, we’re wrapping up a couple of last-minute bugs, and then we’ll flip a switch and turn them on for everybody. So look for them in the coming days.
Russ Savage: 00:13:00.432 And then this is just the beginning. This is our first step towards adding annotations in our platform. We want to support, obviously, ranged annotations, custom colors, support for additional graph types, annotating all the different graph types that our dashboarding tool supports, which is many, so different ways to annotate that data. And then we’ve got some more exciting things coming in that I’ll talk about a little bit later. So really, in summary, InfluxDB Annotations, you can share context quickly on your dashboards across your team of engineers. It’s one way that we enable teams to collaborate on time series data when they’re looking at their operational metrics. Great. So you’ve got your dashboard, you’re looking at your application, you’re investigating anomalies or outages, using annotations, that’s awesome, but we’re a cloud platform for application builders. So how do we support collaboration while you’re actually building on our platform, right? As I mentioned before, our web interface provides a ton of tooling to enable you to create different resources, different queries, different processing scripts, different background processing tasks directly in our web interface, so how do we support collaboration while building on our platform? Anyone who’s used InfluxDB before has known that Data Explorer has been at the core of our query-building interface for a long time, it’s a really powerful tool, we get a ton of great feedback on it, people love it, and so how do we support kind of the next generation of builders building on our platform? And so we’ve created what we’re calling InfluxDB Notebooks.
Russ Savage: 00:14:58.723 So InfluxDB Notebooks are a really powerful way for teams to collaborate and incrementally build resources on the InfluxDB platform. So what is a resource? Resource is anything that you create and save on the platform. So that means that if you need to build visualization queries for a dashboard, you can leverage InfluxDB Notebook to create that information. If you want to build a background processing task to downsampled or enrich your data, you can leverage Notebooks to build that. It’s an interface that allows iterative incremental building of resources on our platform. So this is a screenshot from some data. I recently got a flume water meter, so it’s a sensor that hooks to your water meter, I live in San Francisco, so I was out in the middle of the night opening things in the sidewalk to stick the sensor onto my water meter, but now it pumps in data from the water usage at my house, so I can see how many gallons of water I’m using and optimize. And so I’m leveraging that data here, I’m able to quickly select the data that I’m interested in, and you can see that I’m able to insert a bunch of different cells directly into my building experience, depending on what I want to do. InfluxDB Notebooks were designed to unify the building experience of our platform. And so anyone who’s used our tool before has noticed that - or maybe not noticed, but there’s a lot of different ways to build flux on our platform. They’re all slightly unique, and they have their own little quirks, and so you find some people prefer one experience over another, you end up moving code around manually, and it’s very frustrating.
Russ Savage: 00:16:53.997 So Notebooks exists to unify that experience, they’re designed to allow teams to incrementally build queries or background processing tasks as a team, together, they’re flexible enough, they can guide new users through creating resources without any code, they can unblock power users for writing custom code in our application if that’s what they really, really want to do, and a really, really powerful capability, I mentioned before, you can actually design and build related resources all in a single view, and so if you’ve got a data set that you want to build visualizations for, you’ve got a data set that you want to build downsampling for, you can see visualizations of that data before and after your downsampling on the same page, you can compare, make sure that you’re on the right track, and get that real-time feedback, which is awesome. And then finally, again, you never lose your data. We autosave everything behind the scenes. And so anybody who’s been frustrated with refreshing a browser and losing some form input, that’s not a problem with Notebooks.
Russ Savage: 00:18:05.420 Let’s talk a little bit about a couple of different capabilities that are new in Notebooks that are really, really exciting, for me. So we’ve introduced a new way to get at the metrics and the time series data that you’re interested in, and so we found that the users that are in charge of loading data into our platform really deeply understand and really love the hierarchical structure of our data, right, and so thinking about it in terms of measurements and fields and tags and being able to leverage our data explorer to build out that hierarchy is really awesome. You’ve got that same capability in Notebooks. We’ve got the same interface if that’s what you really love, but we also see a subset of users that don’t necessarily understand all the nuances of the data hierarchy in InfluxDB right away, but they do know the metric that they’re really interested in finding, right? And so I might not know that the field HTTP requests is in a particular measurement, but I do know that that field exists and I want to look at data. And so our metric selector allows you to quickly filter to the metric that you’re interested in, regardless of which measurement it’s in in the platform. And so [inaudible] is really easy to find the data that they’re looking for, the data they want to work with, which is really awesome.
Russ Savage: 00:19:36.735 This is an example of a panel. A notebook is comprised of multiple panels. Panels are kind of like a discrete unit of processing. And so what’s really, really awesome that our development team loves is that they can quickly add and remove and test panels without really impacting anything else. Anybody with a keen eye notice that I’ve got some feature flags set on my cloud account, you may have seen a YouTube panel, so the panels are insanely flexible, you’re able to actually incorporate anything into your Notebook that you want, and so as we go forward and you start to see more development in our Notebooks environment, you’ll see more panels, and we’ll be testing a bunch of new building and editing experiences and getting user feedback, which is really awesome.
Russ Savage: 00:20:32.522 And the second thing I want to highlight about Notebooks is it has a ton of visual components for enabling you to build resources in InfluxDB without jumping directly into the code. But we’re a platform for application builders, application builders love code, it’s great, we want to support that, and so Notebooks are actually flexible enough to support not only the user that isn’t familiar with Flux, a power user that’s really, really excited and wants to customize the Flux however they want, and so this is an example of a coding interface into Notebooks, this is a Flux script panel. You’ll see that I’ve got raw access to the code, it’s got the same autocomplete that Data Explorer has, it’s got the same functions list, so you can see the availability or the capabilities of a Flux, and we add new functions all the time, and those just show up, which is awesome, and then you see real time data preview as you’re iteratively building. And so developers can get real time feedback on their queries and their code as they’re writing it. And that’s one of the reasons people love writing queries in our UI, that real time feedback, Notebooks actually allow you to do more than Data Explorer - you can see multiple visualizations on the same page. So for example, if you’re working with geolocation data, you might want to see the raw data, you might want to see some of the magnitudes of maybe earthquake data as a line graph, but also geolocation data, and so different graph types. You’re able to add those into the same notebook and see them in the same screen, which is really powerful.
Russ Savage: 00:22:27.274 So Notebooks are flexible enough to support many different use cases, many different types of builders building on our platform, and we’re going to continue to add new capabilities and new functionality to Notebooks to really, really turbocharge that developer experience. So the takeaway, if you had to only remember one thing about this section, InfluxDB Notebooks, you can collaborate to incrementally build anything on our platform, and that’s the goal of Notebooks. We’re just getting started, there’s a ton of underlying technology that went into that, and now we’re going to start really iterating, really enhancing the capability to support, building any of the resources on our platform. And as our platform - our platform team is really awesome, and they’re building out new capability and new features every day, and so we want to make sure that we have a UI or UX to build our interface that’s flexible enough to support all that, Notebooks is flexible to support all that, so we’re really excited about it. So those are two different capabilities that we’ve enabled recently in our - oh, in our UI Notebooks available today. You’ll see it in your account, just look on the left navigation under Books, or if you expand, it says Notebooks, and take it out, give it a spin, and let us know what you think.
Russ Savage: 00:24:02.908 So as I mentioned before, so those are two different capabilities that allow developers to collaborate inside our web interface. Again, it’s a powerful web interface that really exists to speed up and enable builders to build applications faster on our platform. But we’re a modern application platform, keyword is modern, and so we want to support modern workflows. And one of the things that has come up in the recent years, with the rise of Kubernetes, has become insanely popular and really, really powerful, is this notion of GitOps supported workflows. We’ve got teams of developers that are building out and maintaining complex, complicated configurations as code, they’re checking that into GitHub iteratively, making changes to that logic, and automatically deploying it to production as it’s ready. And so for those of you who aren’t super familiar with GitOps, this is the textbook definition from Weaveworks. For GitOps, it really describes a path forward for the developer experience. They manage applications based on continuous integration, continuous deployment pipelines, and Git workflows. And so again, it’s how do I always maintain a desired state of my system, how do I tell the system what that desired state is and the system can go about and figure out the best way to reach that desired state and its control plan, right? Again, really powerful in the Kubernetes ecosystem, but we want to support similar capabilities in our platform as well.
Russ Savage: 00:25:57.076 So teams, when they’re collaborating using GitOps principles, we want to apply those principles into InfluxDB Cloud, it’s really easy to do, so we’ve built the system to allow those. So any of the resources that you create, so when I was talking about dashboards, I was talking about background processing tasks, I was talking about any of those queries, you can define those resources as very simple text files, text configuration files, those text files can be checked in and managed anywhere, so a ton of people are checking in that code to Git and building automation there. Our command line tool that comes with our platform, really, really powerful tooling, allows you to quickly apply updates to your resources in your application. What’s really awesome about our command line tooling is it supports configuration profiles for quickly switching between different environments. And so if you’re leveraging developer environment, maybe locally on your laptop, you’ve got a couple of test and staging environments and then of production, you can build out configurations on your command line interface for talking to those different environments and quickly move resources across them. Again, really, really powerful for teams building out DevOps workflows.
Russ Savage: 00:27:27.786 These declarative patterns, they mean we’ve actually built the system so that resources are updated, not destroyed and recreated, and so that can be really powerful. IDs stay the same, and resources are patched as needed or created as needed, the system just figures that out, so you don’t have to worry about it. But you won’t end up with a bunch of different copies of the same dashboard if you keep reapplying the same configuration. And then, again, a combination of our CLI tooling and our APIs mean that you can build really simple scripts that enable powerful continuous integration, continuous deployment. And so the pattern of actually checking your code into GitHub, merging the full request, and then automatically deploying to cloud, a lot of teams are doing that with our InfluxDB resources for their application. Just to hone this in and really make it concrete, just an example of what a GitOps workflow might look like for your team, you’ve got a repo that contains all the different assets that are connected to your InfluxDB account, so this could be dashboards, could be tasks, could be different bucket definitions, anything that you’re able to create in our platform can be saved there as files.
Russ Savage: 00:28:52.643 A developer needs to make a change or needs to make an update to a task that clone that repo locally, they can configure it to work against their local InfluxDB instance, they run one command to install the entire set of resources into their local instance from scratch, they can run a second command from the CLI to actually pull production data directly into their local instance for testing or for validating that things are working, they make their change locally, it has no impact to production, they test it with data that they’ve gathered, everything is working great, the really awesome thing about InfluxDB Cloud is that the open source tooling and the cloud tooling have the same APIs, and so you can leverage the same tools regardless of where you’re deploying your application, which is really cool. They make and they test their changes locally, they commit their code, they open a PR request the same way that you would do with Python or Go code, the developer team collaborates and reviews it, make sure that everything’s working, adds comments, deploy it. Once you deploy it, that code is then picked up by some continuous integration scripting and can automatically deploy that to your cloud account. And so you see new resources updated and added automatically based on the deployment, and then, of course, the really awesome thing about GitOps is you can quickly revert those changes if they have a negative impact, and again, their continuous integration deployment will automatically deploy those changes or automatically revert those changes as needed. So this is a really basic, really simple GitOps workflow, but hopefully, it helps illustrate some of the capabilities that allow teams to actually create and leverage GitOps workflows on our platform.
Russ Savage: 00:30:47.917 Just a recap of some of the technologies that I talked about. Again, there’s a lot out there, it’s too much to cover in just one webinar, but if you need something to do some research on your own and if you want to enable this stuff, our command line interface is an easy way to build scripting, automation, continuous integration against your InfluxDB instance. Part of that CLI tooling includes configuration profiles that allow you to quickly switch between environments: cloud, development, staging, production, whatever you need. InfluxDB templates and stacks are the two technologies that really power a lot of the resources defined as text files. And so a template is the same - the same template that I was talking about earlier, that you’re able to install through our UI, that’s the same technology that powers these GitOps workflows. You’re able to define any resource in the system as a text file and incrementally build and improve that definition and then deploy it using our CLI stacks are a way to manage those templates, and so stacks give you specific stack ID that you can then leverage to continuously update and evolve a particular set of resources in the platform. So again, a ton of underlying technology to enable these workflows for developers to really collaborate on using GitOps best practices.
Russ Savage: 00:32:25.877 InfluxDB Templates - again, big tech slides, which is my favorite here, InfluxDB templates are collaborative GitOps workflows for all of your Influx of resources, which is really, really awesome. So those are the three different ways that InfluxDB Cloud really helps enable you to leverage your time series across your team, really collaborate to build the next-gen time series applications, so looking at operational monitoring and metrics, Annotations allow you to quickly add context to your dashboard so the teams can collaborate on incident investigation or outage investigation or just understanding their data better. InfluxDB Notebooks is a collaborative editing and building experience directly in our web interface, allows you to quickly build out any resource in our platform with - making sure that you’ve always got that context saved or able to jump back in, and multiple team members can collaborate and build together. And then InfluxDB templates, it’s a technology that enables true GitOps-style workflows for development teams collaborating and building large-scale applications on InfluxDB and really, really wanting those GitOps principles of being able to declaratively define your resources and let the system figure it out, and also includes all that rollback capability and continuous integration.
Russ Savage: 00:34:03.106 Really, really awesome set of technology, our dev team is hard at work, and this is really just the beginning of a lot of these technologies. And so really quick before I take questions, I really want to just talk about what’s next and what’s coming. So we’re continuing the collaboration message on InfluxDB Cloud. So when you’re thinking about where Annotations might evolve, I already talked about the next iteration of range Annotations, and what sounds very simple but colors highly-requested capability, so I think the next generation of Annotations would be around automation, and so the ability to overlay event data on top of your metrics in the same view would be really powerful for triaging different capabilities. So imagine you’ve got a stream of deployment events coming into your platform, you can overlay that on top of your network latency graphs so you can see and correlate, “Oh, well, every time I do a deploy, I’ve got a huge spike in network latency, now let’s figure out how to investigate and fix that.” On the Notebooks front, we’re going to continue to unify building experiences across our platform and really focus on making sure that you can build new resources. And so we’re digging into the background processing tasks, next, making sure that you can really, really easily get at the underlying code that’s generated for all of you coders out there, and then we want to focus on building out monitoring and alerting capabilities using Notebooks. So again, ultimately, that’s the area we’re going to be able to build any resource in our platform. And so we want to continue to add that capability.
Russ Savage: 00:35:58.885 And then for GitOps workflows, everything that I described today, it kind of happens outside of our web interface. And so we really want to look for ways to bring those GitOps-style best practices into our interface so you can quickly and easily check in and check out code changes without having to jump into your command line, so we’re looking at different ways and different technologies to support that. All right. So the thing that I really want to leave everyone here, it’s early for me, but I don’t know what time it is where you are. But really the message that I want to leave you with, InfluxDB Cloud, it’s designed for teams that are collaborating on time series applications. So I started this talk with the notion that InfluxDB Cloud supports modern application builders. We know that developers and builders are not building alone, they’re building across teams, and so InfluxDB Cloud, we really want to build and design that - our platform for teams collaborating on building the next generation time series application, and we really want you to get started today. And so I’ll leave you with a huge thank you for joining me on this webinar this morning. If you’re really interested in learning more, jump in at influxdata.com, you’ve got access to sign up for free cloud account today, try out all this capability, it’s really, really awesome. And of course, if you’ve got any feedback when you’re using this, I’m Russ at InfluxData, Product Manager, I’d love to hear about it, just let me know. I’m also in our community Slack channels and pretty much anywhere else you’ll see InfluxData.
Russ Savage: 00:37:50.809 And that’s all I have today. I think the one note that I want to add, as Caitlin mentioned at the beginning of the talk, we’ve got InfluxDays coming up in May, really, really excited, we’ve got a great set of speakers lined up, ton of hands-on training sessions, it’s a virtual experience, you can join from anywhere in the world. This is going to be - I’m going to be joining the sessions in the European time zone, so it’s going to be really, really early for me, but I’m super excited to be involved and hear more about how developers are building on top of our platform. So we hope you join us there. Again, a lot of effort and cool technology are going to be showcased here, so come join us.
Caitlin Croft: 00:38:39.167 Awesome, thank you so much, Russ. That was great. And for anyone who’s interested, I did throw in the Zoom chat the submission form for the Built on InfluxDB Awards. So if either you’ve built something or you know anyone in the community who’s built something really cool and you would like to submit them for a nomination, please, do so. All right, so the first question for you is, can we replace the data pipeline architecture and use InfluxDB instead of using AWS as three buckets for data warehouse and then use Glue and Redshift to query?
Russ Savage: 00:39:19.288 Yeah, a ton of technology buzzwords in that. So what I’d say is InfluxDB is a great place to store and manage your time series data, the pipelines and the tooling and the technology that you use to either bring data into that platform or move data out of that platform are completely customizable and completely up to you, and so we see a lot of our customers are leveraging large scale data pipelines for ingestion. Any time you’ve got distributed systems communicating to each other, it’s really great to have queuing mechanisms in-between there in case of any number of issues; things can get queued up and then reprocessed when everything is back to normal. And so I’d love to hear more about your specific use case, but in general, you can leverage any data processing pipeline to bring data into the system or move data out of it. And then, again, if you want to process that information while it’s inside InfluxDB, you can leverage third-party processing. But we have processing built-in leveraging Flux, and so you don’t need to go to a third party to process that data once it’s inside our platform. The other really cool thing is that Flux supports remote data sources. And so I’m talking about Flux, Flux as our data scripting language that’s available in our platform, it supports the connection to remote data sources. And so if you have data stored in external PostgreSQL instances or different locations, you can work with that data where it is, you don’t have to necessarily ingest it in an Influx first and then work with it, and then you can write that data to wherever you want.
Russ Savage: 00:41:15.226 And so, yeah, there certainly are different techniques and different mechanisms that you can use to process data from remote instances without bringing it into Influx, it all just depends on your use case. So drop into our community channel or reach out. We love to hear more about how we can kind of help that.
Caitlin Croft: 00:41:34.860 Great. Let’s see, so this person has a question about querying and how to write their own queries, “So I must select some of every input data until it shows as a display name on my dashboard. I need to have a query. I searched, but I did not understand where I can write my query for controlling my dashboard.”
Russ Savage: 00:42:02.601 Yeah, so I think there’s a lot of improvements we can make to documentation inflows for this stuff. So if you’re building out a dashboard, you’re building out queries for a dashboard, a great place to start is either in a Notebook or a Data Explorer. And so jump in and you can literally build your query in any of those tools today and then you can export that query to your particular dashboard. And so a lot of users will start in either Data Explorer or Notebooks and bring that query to your dashboard. And then, of course, once you’re on a dashboard, you can edit it there as well. So you need to add new cells or clone existing cells and make changes to it, you can do that all in our dashboard tooling. But really, kind of the starting point for writing queries or writing queries that are backing any of the visualizations in our platform are really going to be either Notebooks or Data Explorer going forward.
Caitlin Croft: 00:43:10.306 How can we engage expertise from InfluxData for reviews on our migration and optimum use of storage? We don’t have expertise in health for InfluxDB or time series, so our team is hesitant to move from AWS S3 to InfluxDB. We want to merge data from different databases as different sources and probably use InfluxDB as a data lake and warehouse.
Russ Savage: 00:43:39.807 Yeah, we’re willing and eager to help you any way we can for those migrations. And so a lot of people, they’ll either reach out directly to one of our contact at InfluxData, reach out to our support at InfluxData, or they can start with just some basic questions on the community forums and get a sense of what that migration might entail. But obviously, we have engineers and solution architects and everybody who does those migrations all the time, we’d love to help you figure out the best way to do it for your use case, so reach out. And worst case, you can obviously shoot an email to me, and I’ll be happy to connect you to the right people to figure out how to bring that data over, so.
Caitlin Croft: 00:44:39.892 All right. Angela, our friend over at the Leiden Observatory said they’re really excited about the new developments, this looks really great, especially Annotations in 2.0 and the Notebooks feature. Can you comment on the status of these amazing features in the OSS version?
Russ Savage: 00:45:00.913 Yeah, great question, and great to hear from you, Angela. I’ve been meaning to reach out and connect to get some feedback on Annotations work, so look for an email in your inbox coming soon. But yeah, both of these capabilities are going to make their way into open source. So next release of open source, for anybody who’s been following that project, we haven’t really continued to iterate our UI there for the past couple of months, we finally figured out and unblocked that capability, so the next release of Influx open source is going to have our latest UI changes in there. I don’t think it’s going to have Annotations and Notebooks because I think there’s some API changes that we need to support as well, but those are both making their way into open source and so they should be available in open source soon. But yeah, Notebooks and Annotations both plan to be an open source project as well.
Caitlin Croft: 00:46:04.433 Cool. If I understand the GitHub’s workflow correctly, it’s focused on making changes to text definition files. If this is correct, do you have any plans to add a mechanism to allow GUI changes to feed into the GitOps workflow, for example, creating a graph as you normally would do through the GUI but at the point of clicking Save, you might be presented with a Commit-like interface that would allow you to write a commit comment that would push the graph config change as commit to a Git repo.
Russ Savage: 00:46:39.424 Yeah, I mean, that’s exactly what we’re investigating. So that’s kind of the next steps that I was talking about a little bit earlier, really want to figure out ways to bring those Git workflows into our web interface. It’s a huge ask. And again, we want to be really smart about how we do it. What you just described is the process and the mechanism that we’re playing around with internally. The same idea of, okay, you have a web interface that’s powered by a Git repo. Once you make local changes, again, you’ve got unstaged commits, you can group those changes together into a commit and push those directly through our web UI without having the jumping out. So yeah, definitely something that we’re looking into, it’s been asked from multiple customers, it sounds like we just found one more user that’s really interested, so yeah, so look for that in the future.
Caitlin Croft: 00:47:36.034 Notebooks are a great way to build analysis combining code, visualization, and text, at Leiden Observatory, we use Jupyter Notebooks for that, with the InfluxDB client, can you comment on advantages and differences on both approaches?
Russ Savage: 00:47:53.273 Yeah, so teams leverage and use Jupyter Notebooks with our technology all the time. I don’t think we’re trying to replace Jupyter Notebook-type experience if you’re using that tooling and loving it. We want you to build in the tools that you really love. Notebooks really provide a foundation and a mechanism for us to quickly iterate and create new building experiences in our platform, it also supports kind of the iterative building that people really love in Notebooks and so the ability to incrementally build up new resources and see the changes as you go using visualizations, you can do that workflow in our Notebooks, we’re looking for more feedback on how to improve, and so if you tested it out and are looking for some more feature capability that either is or isn’t like any of those external Notebooks tools, we’d love to hear about it. The naming is notoriously difficult in the tech community, and so InfluxDB Notebooks, we may have solicited some comparisons that might not be truly valid. But we do want to move in the direction of leveraging in Notebook to iteratively build out data analysis and save that work as you go so that others can jump in and look for it and provide context. And so in that regard, it’s similar to other notebook tools that you’ve used.
Russ Savage: 00:49:42.758 But again, if you’re happy with Jupyter Notebooks, and we have a ton of users that are, continue to leverage that. I would keep an eye on InfluxDB Notebooks as they evolve and see if it’s easier or simpler to do some of the tasks that you’re doing external to our platform there.
Caitlin Croft: 00:50:04.199 Great. Susan also asked if there was a visual example, so I’m not sure if, Russ, maybe you can go back to the slide that you had the examples of the dashboards [inaudible] that. If anyone else wants to [crosstalk] -?
Russ Savage: 00:50:20.906 Yeah, visual example of anything specific or just a visual example of a Notebooks or different capabilities? So yeah, long story short, working with time series data is inherently visual, you can get a ton of insights from simply plotting a data on a line graph, and so we want to bring those visual experiences anywhere that you’re working and writing queries. Notebooks allows you to actually add different visualization types to the same screen as you’re building that out, and so, yeah, anywhere you can write a query in our platform, we want you to be able to visualize that data as you’re writing it, so really bringing that tooling to the users as they’re building applications.
Caitlin Croft: 00:51:09.850 Great. Someone asked if this session is being recorded. Yes, it is being recorded. So the recording as well as the slides will be made available later today. So what’s really cool is, this evening, if you just go and check the registration page for this webinar, it will be converted to the recording, so you can go and share it with your team members or go back and watch it again. Thank you, everyone, for joining today’s webinar. We’ll stay on the line just for another minute or so, just in case you have any more last-minute questions for us. So please, feel free to throw any more questions you have. And Russ is always available in the community Slack. I often see him in there answering questions. So if there’s something that comes up after the fact, you can definitely check there. And I hope you guys all have a good day. I hope to see you all at InfluxDays. Thank you so much for today.
Russ Savage: 00:52:12.001 Thanks, everybody.
Director of Product Management, InfluxData
Russ Savage is the Director, Product Management at InfluxData where he focuses on enabling DevOps for teams using InfluxDB and the TICK Stack. He has a background in computer engineering and has been focused on various aspects of enterprise data for the past 10 years. Russ has previously worked at Cask Data, Elastic, Box, and Amazon. When Russ is not working at InfluxData, he can be seen speeding down the slopes on a pair of skis.