How A3Sec Embeds InfluxDB into Their Monitoring Solution to Provide Customers with Real-Time Analytics

Session date: Sep 22, 2020 08:00am (Pacific Time)

A3Sec is a technology company whose solution provides its customers with better cybersecurity, monitoring and business intelligence. WOCU Monitoring, A3Sec’s unified infrastructure monitoring product, is powered by InfluxDB. Their customers include national courier services, power and utilities companies and large telecommunications. As subject matter experts in big data analysis, they help accelerate businesses by empowering better decision-making. Discover how A3Sec’s WOCU solution enables simple scalable network and infrastructure monitoring, which equips customers to become data-driven.

In this webinar, Alba Ferri Fitó will dive into:

• A3Sec's approach to monitoring by integrating open-source technologies
• WOCU Monitoring's competitive advantage within the industry
• How a time series database strengthens their monitoring solution

Watch the Webinar

Watch the webinar “How A3Sec Embeds InfluxDB into Their Monitoring Solution to Provide Customers with Real-Time Analytics” 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 “How A3Sec Embeds InfluxDB into Their Monitoring Solution to Provide Customers with Real-Time Analytics”. 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.

Speakers:

• • Caitlin Croft: Customer Marketing Manager, InfluxData
  • Alba Ferri Fitó: WOCU Product Manager, A3Sec

Caitlin Croft: 00:00:04.465 Welcome, everyone, to today’s webinar. My name is Caitlin Croft. I am super excited to have Alba here from A3Sec. She is the WOCU product manager, and she’ll be discussing how A3Sec has implemented InfluxDB into their monitoring solution. Please feel free to put any questions you may have for Alba throughout the webinar either in the chat or the Q&A. I will be monitoring both. And without further ado, I’m going to hand it off to Alba.

Alba Ferri Fitó: 00:00:40.610 Hi. Thank you, Caitlin, for your warm presentation. And good morning, good afternoon, and good evening, everyone. So welcome to our webinar. I’ve got to say that I’m super excited to be here, and it’s really a pleasure to be part of your webinar series. Okay. So first, let me introduce myself. Hi, I’m Alba Ferri, and I’m the product manager for WOCU, as you said. I’m focused on marketing and sales by designing the different resources we use to increase our product visibility and adoption like blogs and e-books, presentations, training materials, or webinars like this one. I also help customers discover the value of WOCU and create a vision of how it can be used to solve their problems and help them see how they can get the most out of WOCU. I’m currently working from home as many of you, I guess, with this pandemic situation, but I’m based in Madrid, Spain. I joined A3Sec three years ago when WOCU was just a toddler. And I’ve got to say that this has been a very interesting time in my career so far, getting to know how our product is built almost from scratch. I’ve been working in monitoring for the last six years in different telecommunication enterprises. And I like to define myself as a very curious person. I’m always wanting to understand why things happen. And I guess this mindset fits pretty well with our monitoring field where you’re always looking what happened to your devices, when and why it really happened, and what was the real root cause.

Alba Ferri Fitó: 00:02:39.572 Okay. So now that you know a little bit more about me, let’s head to the agenda. And in this webinar, I’ll be talking about A3Sec’s approach to monitoring by integrating open-source technologies. The second point is going to be WOCU monitoring’s competitive advantage within the industry, and of course how a time series database strengthens our monitoring solution. Okay. So let’s start. I would like to start telling you a little bit about the company that is behind web monitoring, A3Sec. I think having some history about the foundation of A3Sec will help you understand why we ended up creating such solution like WOCU. A3Sec is a spinoff of AlienVault, a Spanish threat intelligence company that was actually acquired by AT&T two years ago. And they had this product called OSSIM, which was an open-source SIEM that was very well known and used by the open-source security community. And, well, those of you who didn’t know about OSSIM, I’ll tell you that it was made up from different open-source integrations like Snort, [inaudible], or Nagios. So A3Sec was founded to provide professional services to AlienVault’s hosting platform for LATAM, basically, for Latin America.

Alba Ferri Fitó: 00:04:19.746 And I haven’t said it, but we have offices in Spain, Mexico, and also Colombia. And some of the people who first founded A3Sec have been involved previously in AlienVault’s OSSIM development and evolution. So you can see that our background is cybersecurity but also, and above all, open-source integrations. So this brings us closer to our first topic. Excuse me. A3Sec’s approach to monitoring by integrating open-source technologies. A3Sec has always been related to big customers due to the fact that they were already using our professional services for cybersecurity. So when Telefonica, which is the former national public telecommunications provider here in Spain, made public an RFI to substitute our monitoring platforms, we took a [inaudible] and sent our proposal. Luckily, we were the chosen vendor. They have very clear pain points that needed an urgent approach for a solution. They wanted to have -

Caitlin Croft: 00:05:36.352 Hey, Alba.

Alba Ferri Fitó: 00:05:38.121 Yes?

Caitlin Croft: 00:05:39.129 Sorry to interrupt. It sounds a little choppy. I’m not sure. Maybe you moved your microphone or something.

Alba Ferri Fitó: 00:05:48.150 Okay.

Caitlin Croft: 00:05:51.007 Sorry. [laughter]

Alba Ferri Fitó: 00:05:52.638 Oh, no. No problem. Okay. And what about now?

Caitlin Croft: 00:05:59.459 Yeah.

Alba Ferri Fitó: 00:05:59.618 Is it better? Yes?

Caitlin Croft: 00:06:01.435 There’s still little crackling but it does sound a little bit better.

Alba Ferri Fitó: 00:06:07.876 Okay. I tried to speak slower. So maybe that helps.

Caitlin Croft: 00:06:15.528 Okay. Perfect. Sorry to break your momentum there.

Alba Ferri Fitó: 00:06:21.819 No, no. No problem. Okay. So they needed this unified customized monitoring tool instead of the variety of forms they had at that moment. They needed to have unified reports because that was actually something super difficult for them at the time because every tool they had was used in different reporting methods and formats. And so they never got right the information. It had to be multi-tenant and multi-user. Because in some cases, they had to give [inaudible] VIP clients or technical account managers that were in charge of several clients, or of course network support engineers that belong to different areas. All these users and [inaudible] needed different views. And, of course, it has to have aggregation capabilities. Telefonica Enterprise has a very unique set of network management centers which have to be able to group together networks from multiple zones. It’s okay, the sound, Caitlin?

Caitlin Croft: 00:07:39.973 Yes. You sound great now.

Alba Ferri Fitó: 00:07:42.556 Okay. Well, and above what I was talking before, they have another problem that was - when Telefonica Enterprise sales wanted to introduce a new service, operations had to reinvent the wheel every time in order to monitor the new service but in the different monitoring platforms. So they spent a lot of money customizing cloud solutions that were really not thought to be used like that. So it was pretty obvious that they needed something much more open to let them simplify their new workflows or vendor integration or services. But there was also another big issue. The tool they were joining with had to be easy to deploy and very scalable. But we want that, right? So when we saw all these requirements in the RFI, we knew there was nothing in the market that was going to fit into Telefonica’s needs. So we started thinking about creating a new solution from scratch made of open-source integrations with a unified [inaudible] just like OSSIM was built. So that’s what we ended up doing. A unified monitoring solution out of open-source integrations.

Alba Ferri Fitó: 00:09:07.182 Let me show you now which applications we choose to be part of WOCU monitoring. We choose Shinken to be our core, the active monitoring engine. It is a Python Nagios Core total rewrite and has [inaudible] flexibility and large environment management. We choose InfluxDB as the best times series database we could choose for storing millions of metrics. And the [inaudible] is based on Elasticsearch. And to join them all, we used Django, the Python-based web framework. Some older parts that are also included into WOCU are collectd, our data collection agent. And integration to OpenStreetMap, our geomap module, which is very useful for massive incidents when you have more than 300 data centers to monitor. Right? We also ship Grafana, the super well-known front end that generates beautiful graphs from Influx time series data. And some others.

Alba Ferri Fitó: 00:10:26.281 Okay. So the result looks like this. All the pieces are unified in only one solution. But even though there are some Spanish words here, I think everyone will be able to understand it right now. And again, to clarify the picture, let me revisit every piece of software [for their corresponding circle. The active monitoring engine. The matrix repository events backend. The inventory module. And the user management module. Those are our core engines which talk to other third-party integrations like NagVis, the map editor, and the ones that I showed you before. And to visualize it like a unified solution and data aggregation, we have our web application made up from React and Django [inaudible] engines. Sorry. So to deploy WOCU, we first create our packages based on the Omnibus project. Omnibus is managed by the Chef [inaudible]. Don’t know if any of you know this project, but we’re very happy someone made it possible. So they define it like easily create full-stack [inaudible] for your project across a variety of platforms. So when we create WOCU packages or flavors, we end up having a common binary that manages the different open-source pieces WOCU is made of.

Alba Ferri Fitó: 00:12:16.575 Here you can see the status output for an all-in-one WOCU flavor installation from a terminal. And here you can see this pretty standard deployment is schema installation of WOCU. It will all depend on how many hosts and services a client has to monitor, how long they want to give measurements [inaudible] or how many users WOCU’s web application’s going to have. You can either use an all-in-one flavor with all these engines in a single server or segregate the different roles into separate servers. So the aggregator is the piece that has WOCU web application. Metric has InfluxDB and the connection to Shinken and the aggregator. The Logger hosts Elasticsearch and the connection to Shinken and also to the aggregator. We call Visor the server that will have the full-monitoring engine and the satellite that will only host the [inaudible] role of the monitoring engine. But having an all-in-one deployment or distributed and [inaudible] deployment will all be very similar to this picture. I mean, this will be a whole server or different servers, but users will access the web app to check the information WOCU aggregates while the different engines do their duties pulling metrics, pushing changes, and executing the scheduled tasks.

Alba Ferri Fitó: 00:13:56.371 If we go even deeper in the picture, here you can see the ports WOCU components use for its internal communication. I will now test for 8086 for Influx, right? And now 9200 for Elasticsearch. And we use port 50,000 to query Shinken status through its Livestatus API. And in the following picture, you can see a look of real client development or deployment of WOCU. In this case, they have two different data centers. And we put three master WOCU servers in each. And they use one server for the metrics and another one for event indexing. And we also added a couple of satellites for it, so they’re group company monitoring. Okay. Let me drink a little bit. Excuse me.

Alba Ferri Fitó: 00:15:11.011 Now that you know our guts pretty well, let’s talk about WOCU monitoring competitive advantage within the industry. We can say that our goal with WOCU is to [inaudible] infrastructure monitoring for those enterprises that have thousands of devices to monitor and want to find a balance between having a very complex business structure and using an easy and customizable monitoring solution. [inaudible]. Sorry. I don’t know where I saved this in my computer. Sorry. And using an easy and customizable monitoring solution. I guess more enterprises are taking care of these because surely after we were fully in production in Telefonica, we started to be included in different RFIs, RFPs of pretty big companies within the country like the IT branch of our national career service provider, some power and energy companies and other large telecommunication enterprises.

Alba Ferri Fitó: 00:16:49.128 And after [inaudible] full business deals, I think we have a pretty reasonable, accurate picture of large infrastructure monitoring challenges. We realize that, in general, companies with large infrastructure or companies not that big, but maybe made up from several small company unions after different buying processes share the same problems. In many cases, IT departments ended up having two, three, or four different monitoring tools to manage because there was no time to do a migration when the company was bought, or they use one solution for vendor X and another monitoring solution for vendor Y because it was included in a deal they closed 10 years ago when their company bought, I don’t know, 100 devices from that vendor, for example. So they set up their own monitoring tool and ship it to IT. Does this sound familiar? There’s a good chance that the company ends up with systems that are old, closed, and expensive, and not connected in between them. But also, even though I really enjoy our monitoring-related stuff, let’s not fool ourselves - and I know that’s not the normal feeling about monitoring.

Alba Ferri Fitó: 00:18:16.222 Okay, we know every company has to have monitoring infrastructure as a base service within IT, but really no one wants to own it. So once the monitoring platform is set up, it won’t change unless it is very, very obvious that it’s no longer working. When this happens, instead of migrate it, they just add a different platform that is supposed to do what the old one does. That makes the process of integrating new technologies or services complex and very slow. Also, if the monitoring solution is proprietary, then we will need to ask the vendor or partner to do the tuning for us. And normally, this costs a lot of money.

Alba Ferri Fitó: 00:19:07.141 And now, what else happens when you have different monitoring tools? The user experience is different in between the solutions. So if there’s a unified view of the platform, there’s a better chance that your people will collaborate, having productive conversations because they all talk the same language. And last but not least, having different monitoring platforms means that you’re working in isolated mode, and we all know what this means. Poor communication, duplicated reports, and worst of all, tired workers. But can you guess what other thing big companies have in common? After all these years, they still have many physical devices from many different physical data center or branch offices with a lot of infrastructure to monitor. And it is great that microservices and containers and telemetry have appeared unseen, but these bigger companies still have a lot of [inaudible] to make profit off. And they are not throwing it away.

Alba Ferri Fitó: 00:20:22.527 So WOCU monitoring’s [inaudible] value proposition is help companies to cut on managing monitoring expense and costs. If you get rid of two, three different [inaudible] licenses and have just have only one type of license managed from a single platform that fulfills your needs, this can make a big difference not only in cost expenses but also in the time engineers spend figuring out in which console they have to look for an alarm. We can do this setup and migration of clients or monitoring solutions, and that is one of that main reasons why enterprises that unify their monitoring platforms [inaudible]. And I guess we are also kind of realizing [inaudible]. Right now, all our engineers have hundreds of hours of monitoring experience behind them, and we know pretty well how to migrate almost any platform to WOCU.

Alba Ferri Fitó: 00:21:26.804 We also have a very experienced development team that can create connectors that would talk to that old [inaudible] or that [inaudible] that was created 15 years ago. And our custom developments are very affordable because we have built WOCU to be very, very easy to integrate with third-party tools and applications because it has a lot of potential. Almost any JSON output can be added to WOCU easily. There’s no need to spend time on complex architecture designs. We tried to be as usable as we can. Always behind getting you to use the fewest possible clicks to get to the relevant information. And now, we will listen to our customers. We know the treasure that is customer feedback. And we give support in Spanish. Maybe this doesn’t sound that important and an issue to me, but there’s still some language gap in our country and maybe in some other Latin American countries. And so some clients really appreciate they can talk in their native language.

Alba Ferri Fitó: 00:22:43.852 So let’s see how a times series database strengthens our monitoring solution. Back in 2015 when we were still arranging the different pieces WOCU was going to have, we found out that there was already a connector between Influx and Shinken. So we set up a small test and it worked like a charm. So we adopted Influx as our time series database for storing all the measurements of the thousands of hosts and services that WOCU was going to monitor. So I guess Influx was in the right place at the right time. It simply uses the [inaudible] feature. You run a query and [inaudible] directly back into the database without round-tripping to the application client. And this feature alone was enough for our processing pipeline time. So the scale tipped to its side. Five years have passed since then, and we do not regret that decision at all.

Alba Ferri Fitó: 00:23:56.030 Now I would like to show you a couple of use cases where our clients use InfluxDB in a great way within WOCU. Our national logistics carrier, Correos, they monitor around 8,000 hosts in almost 85,000 services. They use our monitor solution to know how much traffic of different variety of service they generate. And those metrics are, of course, stored in Influx. This is very important for them because all the traffic is being [inaudible] to different service providers. And they can see all the information in their Grafana dashboard integrated in WOCU grouped by territory, quality of service, or supplier. And every month, they generate [inaudible] records and check if these agreements are being encompassed by their service providers. And so their financial department can then accordingly saving some big money if those SLA’s are not fulfilled.

Alba Ferri Fitó: 00:25:07.313 And the other example comes by the hand of a power supply company, [inaudible], which uses WOCU to monitor 2G and 3G coverage of around 30,000 different devices with WOCU. They want to know how much of the [inaudible] fluctuates because this could end up being a problem of coverage or positioning or something like that. So they can be proactive enough to avoid these kind of situations. But what I’m showing you here is another measurement. Data storage in Influx, their network traffic for device consumption [inaudible], and that allows them to see when their devices with excess usage because when that happens, they know that their signal has been hacked. So nobody likes to be [inaudible].

Alba Ferri Fitó: 00:25:59.416 Okay, so let’s summarize all the benefits with [inaudible] on using InfluxDB within WOCU. The main benefit is resource efficiency. We can achieve a lot with a very small footprint. How efficiently Influx aggregates time series [inaudible], right? At a close second comes InfluxDB retention policies. As the product matures and the infrastructure scales up with [inaudible]. It is great to have an [easing up?] with just retention of data up or down to avoid a catastrophic [inaudible] crisis. And then write speeds are fast. Access to the [inaudible] is super-efficient. You can add measurements and columns as you go, which is very nice. And you only need to create the database, which doesn’t require any skill. So we are very, very happy with Influx for its design features, performance, and reliability.

Alba Ferri Fitó: 00:27:04.394 In summary, we can see how Influx help us in use cases such as quality of service monitoring to meet SLAs and traffic consumption. But it helps us in many other use cases such as an SNMP [inaudible] metric storage, general infrastructure performance, protocol state, WIFI coverage, so, database performance. So I would like to end the webinar talking about the upcoming developments in WOCU. These are [inaudible] topics that we have next. A better UX. As everyone, we’re always trying to be as usable as possible, as I said before. And related to Influx in particular, right now, in our reports module, we have a couple of templates to create different metric reports at a very low level, and those reports use InfluxDB measurement names. And in some cases, it is a too technical approach for our users, so we are translating the names of the measurements that come from Influx to their corresponding server’s name in WOCU.

Alba Ferri Fitó: 00:28:18.962 We also carry out new integrations, and, well, this is also continuous development projects too. And we have new integrations in every release [inaudible]. The next ones are related to network [inaudible], and I’m talking more specifically about Cisco platforms. We’re also migrating to Elasticsearch 7. And we hope by the end of the year that we will have our first phase of NetFlow monitoring module. So that’s what’s going on. And that’s all for now. Thank you very much. Muchas gracias.

Caitlin Croft: 00:29:03.975 Thank you, Alba.

Alba Ferri Fitó: 00:29:04.237 [crosstalk] hear you. You’re [inaudible] for the Influx database.

Caitlin Croft: 00:29:09.743 Perfect. Thank you. So as I mentioned before, of course, InfluxDays this year is going virtual. So we’re super excited to have everyone there over Zoom and Slack. Excuse me. So on October 27th and 28th, we will have our hands-on Flux Training. There is a fee attached to the Flux Training just because we want to maintain a really good student to instructor ratio. So these guys have given this training multiple times. They edit and make it even better every time. So I’m sure it’ll be another great session. And then on November 10th and 11th, we have the actual conference itself. So over the course of the two days, there will be lots of presentations from InfluxDB engineers. Paul Dix, who’s the cofounder of Influx Data, will kick it off, and then you’ll also hear from customers and community members, including companies like Target and Discover Financial. And we also just announced that on November 9th, we will have a - it’s a pre-training. It will focus on InfluxDB as well as Telegraf. So we hope to see you there. Thank you, Alba, for a fantastic presentation. There already are a few questions that have come in, so I’m going to go through them, and you can answer them accordingly. So you mentioned that A3Sec was born from SIEM. Are there any security metrics that you are collecting and presenting to your customers?

Alba Ferri Fitó: 00:30:58.840 We are monitoring, really, the security infrastructure, but we do not [advertise?] any type of security software. What we also do is we send our events for our insights into different [CMs?] or other security software that need our information.

Caitlin Croft: 00:31:31.019 Perfect.

Alba Ferri Fitó: 00:31:31.638 But we do not monitor - or we do not act as a security type of application.

Caitlin Croft: 00:31:40.743 Do you support GNMI? Oh.

Alba Ferri Fitó: 00:31:46.601 I’m sorry?

Caitlin Croft: 00:31:48.331 Oh, sorry about that. Do you support GNMI?

Alba Ferri Fitó: 00:31:56.139 Is that the new Cisco protocol?

Caitlin Croft: 00:31:59.817 Yes, I believe so.

Alba Ferri Fitó: 00:32:01.353 [inaudible]?

Caitlin Croft: 00:32:03.129 Yes.

Alba Ferri Fitó: 00:32:05.888 No, we don’t. Not yet. We would like to, but we haven’t found any customer that still has this technology. As I said, many of our customers have still old infrastructure, and so we serve of what they need. And right now, we haven’t been asked to support this. I hope someone comes with this requirement, but not yet.

Caitlin Croft: 00:32:39.249 It’s funny how there’s so many different tools out there, right? And until a customer needs it, it’s funny how things can get reprioritized. And there is a Telegraf plugin for it, so if you do ever need that, you can definitely use Telegraf to put it into InfluxDB. In the inventory square on your slide, is that for asset management of the customer [inaudible]?

Alba Ferri Fitó: 00:33:17.180 Sorry. Can you repeat the question?

Caitlin Croft: 00:33:22.482 Sure. So on one of your slides, there was the inventory square. And is that specifically for asset management for the customer?

Alba Ferri Fitó: 00:33:36.103 Well, that’s what we use to find out all the inventory, all the assets that we are then monitored afterwards in WOCU. And we have the different automatic tasks that will discover all these assets within the infrastructure. But we do not manage the host [inaudible]. It’s just for the inventory inside a monitoring solution.

Caitlin Croft: 00:34:15.101 Okay. Perfect. So all of this stuff is still hosted by the customer.

Alba Ferri Fitó: 00:34:21.587 Yes.

Caitlin Croft: 00:34:22.645 Okay. Okay. Perfect. Was it challenging for you to implement InfluxDB?

Alba Ferri Fitó: 00:34:35.422 As I said, nothing at all because there was already - in 2015, I mean, five years ago, there was already a connector between our monitoring engine and Influx. And it worked perfect, and that was the main reason why we used Influx. It was an easy decision because everything was working perfect from the first day.

Caitlin Croft: 00:35:01.818 That’s awesome. Were there any hiccups along the way? I know what it’s like to implement any of these sort of tools. There’s always something that you wish had worked a little bit differently or something like that or something that you had known ahead of time.

Alba Ferri Fitó: 00:35:21.967 Well, we’re always struggling with how many data we share in between the different engines. I mean, we have the aggregation capabilities, and there you can group together, I don’t know, clients or levels of criticism or group of services. But do we share this into Influx also in the metrics? Or do we share this into the Elastic [inaudible]? We’ll end up needing this, and sometimes we end up really needing this, and we didn’t have that data. But we have managed it anyway. [laughter]

Caitlin Croft: 00:36:11.765 Yeah, there’s always little hurdles that you have to overcome. How many metrics do you think you’re ingesting per minute? Just kind of curious about the load of metrics that your solution is pulling in.

Alba Ferri Fitó: 00:36:32.480 Oof. I can’t give you an exact number because our clients are so different. We have [inaudible] one client that has 80,000 services, and each service maybe has [inaudible], and another one just one, another one - I would have to study the number. I mean, [inaudible] and seeing how many measurements we store in each case. I can’t tell you the exact number.

Caitlin Croft: 00:37:11.373 It definitely sounds like there’s a great range depending on who your client is, for sure. Yeah. What about for your largest? At the high end, how many metrics do you think? Just a rough estimate.

Alba Ferri Fitó: 00:37:31.543 The largest I think we are storing about two million metrics per every five minutes. I think it’s something like that. Yes.

Caitlin Croft: 00:37:42.327 Oh, wow. Just a few.

Alba Ferri Fitó: 00:37:45.973 But with Telefonica, we use Influx cluster enterprise license.

Caitlin Croft: 00:37:54.166 Right. Yep. When it comes to monitoring, you can collect everything and overwhelm yourself. What advice do you give your customers?

Alba Ferri Fitó: 00:38:09.909 That’s a good one. I always try to make customers realize what they really need. I mean, why do you want to store 15 different database metrics, for example, if you only check 3 of them when the database [inaudible] is in a hurry? You do not need to so many metrics. I try to empower the admins of every device to say which metrics are really relevant to them.

Caitlin Croft: 00:38:52.707 Yeah, absolutely. And how do you help them figure out which is relevant to them? Because I’m sure once you - starting these projects, sometimes it’s hard for them to even know what is relevant to them at the beginning.

Alba Ferri Fitó: 00:39:11.222 Well, it’s not that difficult, but you need to find out the person you have to talk with. I mean, if you go to the DBA person, for example, the admin of the database, and you talk to them, they know for sure which metrics they use. But normally, when you have a monitoring project, you only talk to, I don’t know, the principal, the director of the IT department. That person, of course, does not know about which exact metrics this department, this admin really uses. So I think it should be a broader conversation with all of the users that are going to be using the monitoring platform. You have to talk with everyone that is involved.

Caitlin Croft: 00:40:10.730 Absolutely. Yeah. Absolutely. It’s interesting. For a long time, it feels like the tech industry was all about collecting all the data that you can. And then I think since then, we’ve kind of realized, “Let’s not waste time collecting data that we don’t need or data that people aren’t going to look at. Let’s be smart about our time and only collect the metrics that we need.” I think organizations are getting smarter.

Alba Ferri Fitó: 00:40:35.975 Yeah. Right. But yeah, and that’s where I was talking before in my presentation. I mean, I really love the stuff that is behind monitoring because you can see so many things, but I know a lot of people that doesn’t like it because monitoring makes them really know what their job is. And sometimes when you go with the monitoring requirements, they realize that they don’t really have a good idea of what their job is.

Caitlin Croft: 00:41:12.715 Yeah, absolutely. It’s so interesting. People just get so inundated with the data, then you realize, “Oh, I’m not actually using it.” [laughter] People use the data that is -

Alba Ferri Fitó: 00:41:23.747 Yeah, yeah, yeah, yeah. Totally.

Caitlin Croft: 00:41:25.387 They use the data that they need, right, on a day-to-day basis. As you said, reach out to the people who are actually using the data. Yeah. I know you showed a screenshot of your Grafana dashboards. They’re very nice and clean. Sometimes you can see dashboards that are just so crazy that it’s hard to understand what they’re even trying to say.

Alba Ferri Fitó: 00:41:51.444 Yeah, yeah. That is right.

Caitlin Croft: 00:41:55.622 We actually have one of our users - he calls it data spaghetti when you have a visualization where there’s so much data on the graph that it’s just hard to understand what it says. So he’s definitely another proponent of collect what you need. You don’t need to have everything on the graph.

Alba Ferri Fitó: 00:42:20.487 Yes. Sure.

Caitlin Croft: 00:42:23.497 Thank you so much, Alba. We’ll just stay on, see if there’s any more questions. But it was such a great presentation. I was excited when I talked to you a few months ago, so I’m very happy that we finally got to have this webinar with you. All right, well [crosstalk] -

Alba Ferri Fitó: 00:42:41.756 Yeah. [inaudible]. You know that. We’ve been talking about that.

Caitlin Croft: 00:42:47.910 Yes. It’s funny. We talk and then all of a sudden, oh, it’s webinar time. All right, thank you, everyone, for joining today’s webinar. Once again, it has been recorded, and I will also share - the slides will also be made available online. So please feel free to go back. And if you used the same link that you registered for the webinar, you’ll be able to find the recording later today. And if afterwards you think of any questions for Alba afterwards - I know how it is. You leave a webinar and you think of a question. You should all have my email address, so feel free to email me. I’m happy to pass on your questions to Alba. And I’m also on the community Slack, so if you’re in there poking around. Thank you, everyone. I hope to see you next week at the virtual time series meetup and hope to see you at InfluxDays. Thank you.

[/et_pb_toggle]