Introducing Chronograf Version 1.3
Recorded webinar on Chronograf v1.3—a common administrative UI and visualization experience across the InfluxDB Enterprise and open source platforms to empower developers to build next-generation monitoring, analytics, and IoT applications faster and easier.
Watch the Webinar
Watch the webinar “Chronograf Version Release 1.3” by filling out the form and clicking on the download button on the right. This will open the recording.
Here is an unedited transcript of the webinar “Chronograf version release 1.3”. 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.
• Jack Zampolin: Developer Evangelist, InfluxData
• Tim Hall: VP of Products, InfluxData
Tim Hall 00:01.456 Right. Good morning, good afternoon, good evening depending on where you are joining us from. This is Tim Hall, VP of products here at InfluxData. I’m one of the hosts today and joining me is Jack Zampolin will be walking us through the demonstration. We’ve got our Twitter handles up there if you’d like to tweet at us. We’d love to hear feedback on what you receive today. Can I go through just a bit of a company overview and an overview of the TICK Stack, the technology products that we offer here at InfluxData? We have a number of folks who are joining us today that may not be familiar with the company basics based on the news that we released this morning of the general availability of Chronograf. So we’re going to cover those first, I want to talk through a little bit of the goal of Chronograf, what we’ve set out to accomplish with that piece of the TICK Stack and the technology in general and then we’ll get to Jack with a very exciting demonstration and walk-through of all the new capabilities that are available and then we’ll take a look a little bit about what’s next, where are we going to go from here now that we’ve reached this critical milestone with the project and then we’ll summarize what you saw and we’ll get into the Q&A. So again, if you’ve got questions, pop them into the Q&A panel or into the chat session then we’ll pull those up at the end of the webcast and have a look.
Tim Hall 01:19.012 All right, so first off for those that may not be familiar with InfluxData, the company was founded in 2013 and the focus was really to deliver a modern open source Time Series platform for metrics and events. And some of the guiding principles that we set out to address was developer happiness. Make and easy to use platform that has a minimal set of dependencies that you can install very quickly, build your applications and solutions to address challenges associated with Time Series data and then allowing that to scale out and really focusing on rapid time to value, how quickly could we assemble and build those kinds of solutions. And we wanted to do it all within the open source community, so all of the innovation on feature progression, query language, some other things we’re doing from anomaly detection, data collection, all being done in open source. And the results have been pretty impressive since 2013 we’ve got over 70,000 active servers running the portions of the TICK Stack including the database and over 300 customers around our cloud and enterprise offerings. And in terms of what we’ve done from a product perspective, the TICK Stack is broken up into four key components that address the collection, analysis and monitoring and action for Time Series data metrics and events. And it starts, of course, at the core with InfluxDB, the database itself, for storage and analysis. Telegraf is used as our metrics collection agent that can be distributed and configured. And there’s over 100 different plugins that can be configured to rapidly collect that information and send it back to InfluxDB. Of course, there’s other agentry that’s out there, or you can write custom agents to feed data into InfluxDB itself. These components are loosely coupled.
Tim Hall 03:10.205 We have Kapacitor for monitoring, anomaly detection, scripting, and alerting, that could be done on top of the data. And then, last but not least, what we’re going to talk about today is Chronograf. Chronograf really brings the whole stack together, if you will. It does have dependencies on all these components. Again, you can utilize any or all of these, with Chronograf. But the idea is to provide a single unified user experience, and really accelerate that time to value when you’re using the components in conjunction with each other. And then in terms of clustering security and high availability, these portions, we actually developed in closed source, and make up the part of our Enterprise Edition, our commercial offering. And also are deployed as part of our InfluxDB Cloud, which is our managed software as a service offering that we’ve got for you. But today, again, we want to focus on Chronograf. It’s an exciting day. The team has worked extraordinarily hard over the last six months on this. And just taking a step back, some of the goals of Chronograf. We wanted to provide a unified experience for the TICK Stack. This includes administrative capabilities across the different components, in terms of setting up things like database creation, doing user management, and most excitingly, bringing together the alert configuration for Kapacitor; bring it out of the scripting realm, and into a user experience that can be delivered through the browser.
Tim Hall 04:35.808 Second, we wanted to deeply—we wanted to provide a deeply integrated experience for data exploration. What this means is, allowing you to rapidly create and visualize queries using the capabilities within the database itself. So for example, within InfluxDB, we have the notion of measurements, fields, and tags. And in some context, those elements need to be visualized, shown that you can use them as selection criteria as you’re building your queries. And we believe we’re best positioned to surface that information to you from a browser perspective, and allow you to dive in and leverage the way in which we have stored that data. Second, when you’re using Telegraf, we have the ability for you to expose the pre-built dashboards based on the metrics that Telegraf is sending into the database. What we believe this allows you to do is deliver rapid time to value, meaning you don’t have to sit down. And the first thing you do with Chronograf is you have to write a whole series of queries. If you’re using Telegraf to pipe those data elements in from the various sources, we know what those metrics are. We’ve pre-built some dashboards for you, and you can immediately start exploring that data. And of course, you’re not locked into our visualization. You can build custom dashboards using these collected metrics and events. And of course, we’re delivering all of this in open source. So in terms of actually seeing and not just talking about what we built, I want to turn over to Jack so we can do a demonstration on Chronograf and the latest edition.
Jack Zampolin 06:18.168 Okay. So Chronograf, as Tim said, is going to be a commented administrative UI and visualization experience for the full InfluxData Stack, both Open Source and InfluxDB Enterprise. It’s going to make it extremely easy to get started and get some visualization very quickly just by getting data into Influx. So let me show you how easy it is to get started with Chronograf now. I’m using InfluxData sandbox which is just the full stack all connected together with a little Docker compose file. And we’ll go through the installation here. So the first thing you need to do is connect your InfluxDB instance to Chronograf. Here I’ve got it running at local host 8086. All right. One thing to note is that Chronograf supports Oauth so we’ve got a number of GitHub, Heroku, Okta, as well as generic Oauth providers so you can plug it in with other single sign-on solutions. Once you’ve configured Influx, all you need to do is configure Kapacitor. And you would configure your Kapacitor alerts so if you want to connect to Slack or any of the other alert outputs we’ll talk about this section later. In the theme of being a management console for the whole TICK Stack, we’ve got database management and user management here. So database is very easy and then adding the retention policies. Retention policies tell you how long data should last in Influx. For example, you’ve got some data that you don’t really want to keep around too long, maybe two days. We would keep it in that foo bar retention policy.
Jack Zampolin 08:27.139 User management is another nice feature here, so creating users is a quick task here and then just giving them permissions. If you’re connected to an InfluxDB Enterprise cluster here they’ll be some more fine grade permissions available. There’s also a visualization of what queries are running on your database, so if your queries are taking a long time, you can easily pop in there and kill them. So I got a InfluxDB instance running that’s got a bit more data in it. I mentioned earlier the Oauth. This is what that looks like. I’ve implemented GitHub here. One thing Chronograf can easily do is explore your data. So if you’re familiar with InfluxDB, there’s databases, measurements are much like tables in a sequel database. When you drill down, you’ll see the tags down here as well. So this is identifying metadata that would explain where the data is coming from. So hosts, for example, we got a number of hosts on this cluster. And as I’m building the query down here, it’s going to generate the sequel-like query language that InfluxDB uses. And this is very easy to edit. So for the last two hours, upload verifications and easy visualizations there. And if you need to change time range, you can do that up here. So if you want to run it for the last day. Also, the auto-refresh interval is up here. If you’re using multiple InfluxDBs, this is going to indicate which one you’re connected to. And then, there’s some clicks for the toolkit. So if you want to focus on a certain section of the ground, zoom back out, anything like that. One of the really nice features of Chronograf is if you’re running Telegraf, our collector, we know what data looks like. And we’re going to give you some pre-can dashboards by default. For example, this installation is monitoring Kubernetes cluster. And we’ve got a number of different services running on top of that. So things like memory and CPU for various services, network ingress and egress, passwords that you build yourself already will give you those for free. And then, we do a pre-can dashboards for about 27 of the different Telegraf plug-ins. For example, InfluxDB here gives you a nice view of how the InfluxDB installation is performing, how many series, how quickly data is being written, common things like that.
Jack Zampolin 11:36.311 Chronograf also has dashboarding. Creating dashboards is very quick. Just go edit yourselves and add a query. I’ll show you that sample CPU query I did earlier. And this is the exact same query editor that you’re going to get in Data Explorer. And in this same way you can edit this query up here, so if we want to use a different aggregation here for the max usage user, you can change that right from the console. Also of note, there is a number of different visualization type, maybe you want this in a stack one graph, step-flat, a single step is not necessarily applicable for this one. But we also have line in single set as well. Another nice feature in Chronograf is template variable section. So for example if you have a host-level deck, or if it’s communal across all of your hosts you might want to have the little drop down to switch hosts very quickly. So we will call this variable host. This is via tag values. We just want to see all the different host values that commonly come into here. All right, that looks good. Let’s go ahead modify that cell to work with that host variable. You’ll see down there to what variables popped up. Chronograf will also service any error codes that are coming from the server on the bottom.
Jack Zampolin 14:14.434 Moving on. Another really nice feature from Chronograf is the Kapacitor alerting. So if you’ve used Kapacitor before you would need to write Tick script to do that alerting. Chronograf give you an easy way to generate those with a visual UI filter here. So creating a simple CPU alert, very similar to working with the Data Explorer. You just drill down through your data, select for tags and fields. You will see the query get build up here, so you know what data set you are working on. In setting thresholds down here is extremely quick, maybe even a higher threshold. We also have relative alerting so if you want to [inaudible] moving average is standard deviations beyond where it is supposed to be, or alerting on the absence of the matrix so one of your host reporting, you want to alert. As we saw earlier, when we are setting up Chronograf, there is a number of different outputs that we support. There is an example down here. And setting the content of the message, you have got a couple of template variable that comes from a task you are running. One of the really nice ones is HTTP. So if you are looking to hook in some existing automation system, this will send an arbitrary post request. Json, down here, is using the different template variables. Chronograf also exposes a list of the recent alerts that you can filter by—here I’ve got host is coming out as one of the values. So I’m looking for a specific host, and you can filter for all of them. Set the time range with the convenient date checker. And that’s all the big features from new Chronograf. So with that, I’m going to go ahead and kick it back over to Tim.
Tim Hall 16:26.056 Thanks, Jack.
Jack Zampolin 16:28.063 Yeah.
Tim Hall 16:29.289 Really appreciate you walking through all the great work that the engineering team has done over the last six months. We started this journey back in November of 2016 with an original blog post from Paul Dix, our founder, to really deliver all of these kinds of capabilities in open source, the team has worked very hard. Just driven to—these bi-weekly [inaudible] where we delivered updates of new features, bug fixes, and capabilities over the last six months. And with all the things that you’ve seen, the question is of course, what’s next? And we’ve got some questions in the chat on that. Number one, folks are having a little bit trouble with the download link for Mac. We’ll go ahead and get that fixed up right here after the webcast is over. Second question we had is on millisecond time range resolution. If you want to monitor progress of that one today, it’s not available with the GA today but it’s issue number 1449 of our tracking community. And we’ll go ahead and work on that over the next series of releases. In addition, there’s other visualization options that we, of course, would like to deliver, from gauges and bar/pie charts, to improvement in the delivery of information through tables. And one of the more popular request that we’ve been receiving and discussing internally is the ability to do time shift effectively. Folks want to analyze different time series, different data sets that are really focused on human understandable times. So for example, if you wanted to compare a Monday to Friday time range of a whole set of series on top, or a quarterly time range from quarter to quarter, can you do that and visualize that? So we’ll look to address those kinds of queries and visualizations of that information as we continue on.
Tim Hall 18:27.328 Next, we also want to deliver the ability for you to do configurable views by application type. So today, if you’re using Telegraf to feed in the metrics into InfluxDB, Chronograf will actually show you the applications that have been configured. Effectively, the plug-ins, that Telegraf has been configured to send those metrics in through. And what we’d like to be able to do is allow you to build a consolidated view that potentially cuts across the various layers of the stack. So you could have infrastructure, database metrics, app server metrics, queuing statistics that all represent at a particular application that’s deployed within your infrastructure. We’re also looking at delivering a new landing page, so effectively today when Jack logged into Chronograf, you land on the host’s page. And that host’s page shows you all of the data that’s arriving from Telegraf broken down by individual hosts. We’ve gotten a number of requests to have different styles and information that’s available on the landing page, and this is also something that we’re working on as part of us preparing Chronograf to be delivered as part of our InfluxDB Cloud offering. So if you have suggestions about things you’d like to see as prominent on a landing page, we’d love to hear that within the community. And then beyond that, where do we go from here?
Tim Hall 19:49.406 So we did a great first step in terms of surfacing up the ability for you to configure alerts within Kapacitor through that wizard that Jack showed, and the question is what are the next kinds of things that we want to enable through the UI to allow you to take advantage of the full capabilities of Kapacitor. And one of the things, again, that’s a very popular request is the ability to rapidly assemble a downsampling script for Kapacitor. And effectively, for many types of monitoring data, the value of that data goes down over time, meaning the precision by which you need to store, analyze, and leverage that data decreases with time. And so what folks end up typically doing is downsampling their Time Series data to reduce the amount of precision that they have from, let’s say, if you were storing metrics every minute, you may want to store them by hour or by day or by week, again, depending on how far away in time you get from the original capture of that information. And so that whole process is called downsampling. Kapacitor provides a mechanism for you to be able to do that very rapidly. But, of course, we’d like to ease the process of you actually defining those scripts and being able to debug and deploy them very effectively. And so related to that, we’re also looking and contemplating to build a full TICK Script editor which would include log visualization. So as you’re going through the process of building and developing your TICK Scripts, if there are errors that are associated to it either design time, where the TICK Script editor could catch, or run time, where there are messages being piped out to a log file, we’d like to be able to surface that information to ease the entire life cycle of the development of those scripts. And so look forward to those modifications and improvements as we continue down the path.
Tim Hall 21:49.806 So again, just to recap, we’re really excited today to finally launch Chronograf 1.3 as that integrated user experience for the TICK Stack across our administrative use cases that Jack walked through: exploring the data, building your own dashboards, and, of course, defining alerts to leverage the capabilities that we offer in Kapacitor. It utilizes both the InfluxDB from an open source perspective, and you can use the same binary to access the Enterprise edition for those of you who are Enterprise edition customers. There’s nothing additional that you need to consume or buy. It’s available and works across both editions. As part of this work, we will be removing the admin interface as part of the 1.3 release of InfluxDB. We announced this deprecation earlier, and hopefully, you understand we’ve really focused and put our Engineering wood behind the arrow here to deliver a much more compelling user experience than we ever had with that app and interface. Second, because Chronograf is Enterprise edition aware we will also be retiring the Enterprise web console. So for those of you that have been using the Enterprise web console, we’ll announce its deprecation here as part of the 1.3 release, and then sunset that capability as well. One of the things I have to tell you from working with a number of customers through the support channel is that the general setup and installation of Enterprise Web has certainly been challenging for folks and I’ve had really none of the same challenges with our experience with Chronograf. It’s very simple to set up and use and connect to, either the open source edition or the Enterprise edition and I’m very thrilled to be able to offer that to customers. It’s a much, must faster kind of value and a much richer user experience.
Tim Hall 23:41.930 And then last but not least, as part of InfluxDB Cloud we’re going to be working very hard to deliver Chronograf as the common user experience for all of our InfluxDB Cloud customers this summer. And effectively, what we’ll be delivering is more than a managed databases service. So effectively today folks have been leveraging InfluxDB and using other visualization tools and building applications on top. What we’ll be delivering through InfluxDB Cloud by default will be Chronograf as the administrative and exploration tool for the cloud infrastructure, along with the ability for you to configure alerts within Kapacitor as part of every subscription. And of course, you can expand your subscription and upgrade to full editions of Kapacitor and gain access to its full capabilities as part of InfluxDB Cloud. But we do want to drive folks to using Chronograf by default as you explore take advantage of all the capabilities we have to offer from InfluxData. Of course, we got a bunch of additional resources for you. You can download Chronograf here off of the Influx website. And again, we’ll be fixing the link for Mac, for those who’ve been waiting and asking for that, as soon as we wrap up the webcast. We have the documentation up and available as well for Chronograf, this new edition and all the features that Jack walked through. There’s a fantastic getting started guide that teach you through all of the steps and functions to effectively use this great visualization tool.
Tim Hall 25:21.905 And then last but not least, we encourage everyone to engage with us at community.influxdata.com. This is our branded and hosted instance of Community. We had seen lots of groups sort of cropping up and starting in Google groups, and stack overflow, and other places. And what we thought what we would do is really provide a first-class experience for folks that are using our tools and technologies to collaborate with each other, collaborate with us from an engineering support, product management perspective. So please check out the Community. It’s free to join. If you want to post questions, you do need to register. But we’re looking forward to continuing to build our relationship with all of you form a community perspective, take your feedback. And we’d love to hear your experiences of using Chronograf. So with that, we’re going to move to some Q&A, and we’ll see if I can…
Tim Hall 26:32.846 So let’s see, we’ve got some questions here. So the first question was about, again, visibility of a millisec of time series. Again, track that within community. 1494, I think is the issue number for Chronograf. Check that one out. Scatter plots. So typically, when folks have been creating visualizations for Time Series data, the X-axis is almost universally time. And so it’ll be interesting to kind of dig in further in terms of what folks want to do with a scatter plot. I know we have talked about that internally, but we’d be happy to engage with some of you that have that requirement going forward. If you have other questions, go ahead and drop them in the Q&A section or the chat section, and we will be happy to answer them. Okay. Well, it looks very light on the chat. So I hope everybody enjoyed the demonstration from Jack. I want to thank Jack for preparing the demo, and walking through it so eloquently. Appreciate everybody’s participation. Oh, we got one more question coming out from someone talking about LDAP Integration. So great question on LDAP, and one of the things that we’ve seen over time is LDAP is sort of a historical credential store for a lot of security infrastructure. One question I would have you ask is whether Oauth is in your future as supposed to LDAP sort of being in the past. As we’ve been exploring this feature with various customers, we have uncovered and discovered that there’s lots of Oauth usage sort of cropping up so whether it’s Google or whether it’s Okta for Single Sign-on, you might be surprised within your organization that an Oauth provide exists that you might be able to integrate with. And also you might want to check with your LDAP provider, they also make have Oauth capabilities native now. So for example, I believe the Windows Active Directory may also have an Oauth provider that you can turn on and activate but check the documentation. We know for sure the Azure environment, if you were to deploy in the Microsoft Cloud, Azure provides Oauth capabilities that you could absolutely connect to.
Tim Hall 29:12.896 Another question is about Vocalerts and if you say you’re implementing PagerDuty, absolutely you can quickly integrate with a whole wide variety of alert targets including OpsGenie, PagerDuty, Slack. And then there’s generic outputs as well that you can direct to HTTP files, and so on, and so forth. We happen to actually use PagerDuty here, and heavily utilize both Kapacitor, and PagerDuty to alert us when there are any issues or things that crop up with InfluxDB Cloud. So yeah, that’s absolutely possible, and it’s super easy to configure those alert targets. The dashboards are customizable per user so each instance, in terms of the user login, you can affect and change those dashboards on the user basis. Good question. So next question is about packaging, “Are there plans to provide development packages for the ARM Builds so Raspbian or Raspberry Pi?” I believe we have the ARM package already available for the other components of the stack so it’s shouldn’t be a big deal for us provide them for the ARM platform as well but we will definitely check on that, and if they’ll be available for the downloads page as well.
Tim Hall 31:04.157 Let’s see, any other questions? All right, I think we got to them all. So again, I want to thank everybody for joining today. We’re super excited to provide a Chronograf as a generally available part of TICK Stack today. Definitely leverage it with the open source edition, or Enterprise edition customers—we look forward to you using that within your installations. And I want to thank everybody for joining, we’ll hopefully see you in the Community.