How to maximize your deployment value in Splunk AppDynamics

I good morning and good enough, good afternoon to you all. Once again, I am Sanjeev. I am one of the customer success managers and app dynamics based on.00:00Of Mumbai, India, and I have with me Kevin Whoe, who is based out of Sydney Australia. He is one of the expert CSS that we have on the panel today and today in our.00:20Expert inside session, we will be discussing how to maximize your deployment value in Splunk appdynamics. So, let's start Kevin, if you would like to introduce yourself before we begin, please go ahead. Yeah, sure.00:34For those who has been attending previously, you can see me at a delivering the session or lurking behind the scene, answering questions, holding the meetings, recording and stuff. Yes, I'm basing in Sydney, Australia. My basic role.00:50In Cisco AppDynamics or Splunk AppDynamics is basically managing customers in Australia, New Zealand. Sometimes in the other APJC area, but this expert insight series is one of the.01:10Project that I have with you guys to deliver different topics from to observability cloud from splunk. So just very quickly I'm I had some really good suggestions just now from Pradeep about having a analytic sessions.01:26Feel free to throw me these ideas. I'm happy to talk to the team and have someone delivered as you wish and we're open to a lot of suggestions and yeah, thanks for coming.01:45Yeah, and just for your information, you can ask your questions and like Kevin said, you can provide any suggestions that you have in the chat. We can address those questions or sessions maybe during the, the presentation or towards the end.01:59You can address them and answer your questions.02:19Yeah Kevin, I handed over to you. Yeah, so today's session is an interesting one. It's out of the foundation series as well as the advanced series. This is pulled out of different topics to.02:22Basically trying to build it from a ground up but from other resources truck and giving you.02:42Some of the users that has been using AppDynamics for a little while, it might have become a bit stell on some of your work or you're trying to optimize your deployment and think of what can you do or you're stuck at somewhere. Try to give you some pointer.02:50I'm trying to provide some of our infield experience customer feedback and give you some pointers either from the appdynamics principle standards or from our experience with customer, what they have found, what we.03:10You think it's helpful. So just to begin with, you, you, if you still remember when you start off AppDynamics.03:30Out of the box it comes with a lot of tools, it's very useful for the auto instrumentation in many cases built a lot of business transaction and their four flow maps with you guys. Very useful.03:38You can see the transaction and each business transaction for the technical folks can really understand and it comes up very easily. Each of these pages on the UI are dashboard orientated, so we try to give you what we think is important.03:55So out of the box is very helpful. We are trying to have this wide and depth approach where we get as many things as possible and trying to show you the right metrics or the right logs in the analytics side. However, most of our customer that struggles.04:15To have the typical alert storm or they spend a lot of time to clean up and then build custom work. That ends up delaying the ROI and when they're renewing, they're like, I don't see much value out of it. Why is why I am having detections every 2 s for those who use their mobile.04:35I also didn't like it as well. So.04:55We want to trying to put a lens on, ok, why is this happening? What sort of principle should we start or what should what a framework should it be when we are using appdynamics? How do we utilize its strength? Appdynamics strengths is always around the business contexts translating from.04:58What data that you have ingested. So we were trying to remember the alert and response model we have all these framework. In this session, I want to give you a framework of visibility, insight, and action, to trying to.05:18Use this lens to understand, ok.05:39When it comes to visualizing your data, how should it be? What is it, what, what's the right call when it comes to getting the best value? When visibility becomes inside, which is meet visib, which is metrics or information that's meaningful to you, how should it be?05:42What sort of tool set do you have and eventually obviously actions.06:03That is what we're gonna talk about today. Obviously, without talking about integration, there is NO additional value that we can't, we can't just say how to maximize your value. Integration with part of it. And if time allows, we will go through some of the things that we talk about here on a demo.06:08Controller but yeah, let's kick start.06:28Sanjeev if maybe you can talk about the introduction of visibility perhaps. Sure Kevin, thank you. So AppDynamics has a set of features, amazing features that help us have the visibility over our apps that we are going to monitor and.06:34The whole application environment. So talking about those appdynamic AppDynamics automatically discovers and maps your application environment. This creates a visual dashboard, you know, that provides a clear overview of your entire system. You don't have to manually set up.06:52Set this up and it is automatically provided. Then we have monitoring of network and servers, so it tracks AppDynamics tracks performance and health metrics across various components. That may include your network devices, your physical and virtual.07:13Servers, any environments or, you know, actual end user experience, which we also call as digital experience. We also have full stack observability. This is across applications. So, what, what this means is AppDynamics provides.07:32Comprehensive visibility into, every layer that could be from the application code to the underlying infrastructure and how users are interacting with these applications. We have real time infrastructure health and performance monitoring, so.07:52The infrastructure components are continuously monitored in real time to detect issues like any server failures. It could be a CPU or memory spike or network problems that could impact the performance of your applications.08:12AppDynamics AppDynamics agents, they also collect data, different appdynamics agents will collect different types of data and the flow maps use that data to correlate and visualize it.08:29Which shows how components interact and if there is any bottleneck of failures that are occurring, that will be shown in a visualized manner. Let's try to understand these.08:44How the application works on these features. So Kevin over to you again. Yeah. So speaking of flow map, this flow map is actually from one of my customers.09:00This is actually something that I picked up. It's not a new customer, but a customer that has been taken care of by somebody else perhaps a partner and now falls onto my lap and in a way.09:18This gives you a lot of visibility. That's enough visibility, but it's useless for the customer this is the biggest problem and the amount of cleanup work doesn't really worth.09:34The value that they that that they pick up. That's what's happening with this customer and they don't know what to do with it. They don't know, what sort of things they can manage, they can monitor, but at the same time, they also know everything is there.09:51When they trying to search for a business transaction through perhaps an URL, they can find it. When they're looking for a metric, they can find it. But yet, the most valuable from my, like from my perspective, the most valuable graph visualization, which is the flow map, it's not working for them.10:10So from here the implication is that if you have that visibility, you have that much information ingested. It doesn't really mean that you have what you want or it's visualized enough for you to understand it, at least from the eyes of the human.10:29So we need to ask a question to ourselves, what is the actual KPI? What is the actual KPI for yourself or your team, for the organization? How does it align with each other? From there, you will want to think of, ok.10:46If that's my KPI, say I want to build a dashboard for my boss and these are the things you want to see. Why do I need that much data? I only need that, those metrics for that dashboard and those JSON files to build that UI.11:03So, one of the things we have always talked about is you want to have a few small success when you start your journey. In terms of the flow map, you want to make sure you have.11:20Some tiers and notes that are most important, really prioritized and same as business transactions say we always want to always want to start with five or 15 depending on your on your team size and all that and then make sure these are very important to you the most prioritized.11:38Eyes in terms of project, in terms of your sprint or your long term view, and then make sure they are correct, make sure they have the right grouping.11:58And it's against the right baseline that you want. Is it a, year to year seasonal or just every 15 days baseline? If you have a KPI on a hard threshold, you can set that as your.12:08As your baseline, that's fine, but you need something. And then from there you want to translate into business context, basically your business transaction, you don't want it to be in URL, you want them to be proper. Once you have these set up, you want to lock down your environment. So everything else that might.12:25Would be useful but not now moved to the other bucket.12:45BT lockdown allows you to focus on what's important to you and then you slowly ingest the rest. If you want to if you do another another another way around, you will probably just end up with this. There is NO way around it.12:50And to optimize business transaction, there's other I'm sure like people who are familiar with AppDynamics will know that you want to have your segment set up properly so that it doesn't split into.13:08Many's but same function, this is a transaction. These things are also very, very helpful, but the the idea is that you want to set yourself up to success, not necessarily be like, ok, in three months I can, in three months I can monitor everything.13:21But maybe in two weeks I have one thing that I have set up properly, I understand what it is, I have the alerts, I have the right response, everyone knows what's happening. These are the small success thing we want to see, and then going forward slowly build them up and train your colleagues to do the same thing.13:39That kind of multiplies the ROI for each of the environment. This is a bit of a, I would say a preview of the next EI session in the future simply because we are running one.13:58In 29th of September for the smart agent.14:18Now, just now we talk about when you're using it like cleanup and all that. The other issues that many used to have is the what we call the fleet management or how do you manage the agents themselves, that directly affect how you manage your metrics.14:22Or patching in terms of security, smart agent comes in as an agent of agents where, for those who have used them, will have some form of auto detection to understand the environment.14:41It deploys on an ongoing basis in the UI. You can just go and go upgrade rollback, and you were in that particular smart agent page, you can also understand what's updated, what version it is, have you rolled back, things like that. Smart agent also understand itself, so.14:57It will have its own rollout, own update as well, and it has a CLI interface, a CLI environment for you to run it if that's what you want and you can do it remotely.15:17That's basically if you don't have like a good dev OPs pipeline set up for AppDynamics, this is a very good alternative, and in terms of giving you visibility and just to kick start to get more value out of it from your time.15:30Smart agent is really well used in bigger environment and if you have a very rapid devops cycle or a very strict security environment that asks you or force you to do upgrade for any.15:50CVs. This is a great choice. I won't do too much about smart agent. It deserve its own session and again it will be delivered on 29th of September. Now we'll talk about insights.16:06Anything from you Sanjeev? Sure, thank you Kevin. So, I hope the screenshots of the flow map and the other details they they helped you understand how visibility works and how how do we get visibility over our.16:24Application environment. Now let's look at some features that provide us insight on those on the on the data that we have through visibility. Let's see how that provides some business related insights on that. That helps us monitor apps better and.16:43Tackle any issues that we see. Could you move to the next slide, Kevin? Yep. Thank you. All right, so AppDynamics it links application performance metrics with the business outcome and user experience data, which helps.17:03Us understand how technical issues affect our business KPIs. Then we have artificial intelligence or machine learning which can be used. The system can use it to automatically detect unusual behavior or performance related.17:24Anomalies and helps.17:44To pinpoint any root cause analysis or root cause quickly and you know as less time as possible with the help of the AI and ML. Then, AppDynamics also establishes normal performance baseline. So we can create a baseline of.17:46Our own and then according to that these baseline for the metrics that triggers alerts the alerts can be set after the baseline is created. And because of that the trigger the alerts will be triggered only when there is a deviation.18:06And all this reduces false alarms and, you know, provides more time or more focus or, you know, focus attention on the real issues. So if we want to omit any.18:26Alerts. There are up to a certain extent on any, e.g., any CPU spike or memory spike or network related issues, so there are not considered any real issues. We can create our baselines accordingly and then.18:41Have the actual alerts let us know that there is an issue that we need to provide or focus on. Then key performance indicators are analyzed alongside events to identify patterns and correlations that that you know might explain our performance related problems that impacts the business.18:57AppDynamics also, I mean it it not only finds technical bottlenecks, we have also you know AppDynamics provide prioritize it prioritizes these bottlenecks.19:19Based on the impact on business processes, this enables the team to focus on the matters the, you know, the issues which matter the most. So once again, let's try to understand these features in more detail.19:33And how they applied with Kevin's expertise. Kevin, over to you again? Yeah, so insight is basically trying to translate visit what's visible into, insight that's relevant to you and often.19:52These insights are.20:12Yes, you can use it as a monitoring, but often we want it to become something on demand or proactive. That's where alert and response model comes in. Few of you who has experienced will know that once you get into the alert and response page.20:14You'll see this particular image on the right. It's quite straightforward. What we are trying to do is.20:34Okay, you have the visibility, you want to use this into, to give you some insight, but these insights shouldn't be, can be available on demand, but at the same time you want this to become proactive to tell you when something happens instead of.20:45Nothing happened, nothing happened, it's just keep bugging you that way or you go there and check it every now and then if that's your job, that's fine.21:05But ultimately we wanted to put it into this other response model, it translates these feasibility into some form of responses in a way when there's outages happening, it provides an extra steps for you to.21:13To notify the relevant team to get additional information through snapshots.21:32And, if sometimes it can become actionable insight for e.g. it gives you a threat dump if resources is allowed or in the right programming languages and additionally we have this AI ops that's.21:38For those of familiar has recently released on.21:58Without you manually putting some input, we can give you some output just by understanding the metrics themselves through correlation and machine learning. And then obviously you have that RCA that would cost analysis at the end to wrap it up.22:03In this sense, when we're talking about maximizing maximizing the value is really on number one you want these health rules to be set up in a way that's repeatable, it's meaningful, which basically means that you want to have as many.22:19Metrics or threshold that's available, but yet it is sensitive enough to trigger you in the right time. In a previous session we talk about the levels of proactive monitoring where.22:41The earlier you you monitor, the earlier you are being proactive, the less accurate it becomes, but then when you expand the metrics, the amount of metrics that you ingest, the slower you get and yet you can accurately get the actual issue, it's it's just both end of the spec.22:54So, the arts of setting these up are really basing on, ok, how do you, how much do you know about your environment? How much information do you have about the organization's events, its activities, and.23:15How much data you have from previous history so that it predicts the future. It's basically a very mathematical zero. Now if it's too difficult, we obviously have this anomaly detection that is like a, I wouldn't say it completely replace this model, but a good crutch.23:31To, when you have, oh, I all these all these health rule and all that is working, but then I still think something might be missing, then this is where it comes in or you want to have some early success in your environment in the early days.23:51These are something you can set it up without understanding much. So for our anomaly detection, we don't just understand what metrics and it's like, ok, past history, and then these are the spikes, I will just repeat it. It runs correlation through.24:11To multiple metrics and it's it try to understand the relationship between each other and what does it constitute an issue or an event. So, oh, it seems like I've lost a few of my slides, but that's ok.24:31So for the business transaction we use course per minute and we also use error rate to correlate. When it comes to database, we will use number of connection and also time spent in execution.24:50Plus calls per minute to trying to understand, ok, will these three happening at this certain dynamically, then this might this might be an event. And our detection has expanded from business transaction to database to almost every single section of monitor.25:08In appdynamics. They use different metrics metrics so it will be very fun and useful to.25:28Have a, at least having a goal. In anomaly detection, you can toggle its sensitivity. My experience personally with the customers medium is the minimum low basically doesn't do much.25:39You want to have something happening and then you kind of understand anomaly detection, how does it work? And, is it working for your environment? It takes time for the model to train itself as well, so once you toggle it on.25:55You will need to wait for it to have this training completed signed, then it will start to generate event. Everything is automated. You don't need to set up any health rules if you don't want to, obviously if you want to, you can. You, it will have a certain amount of health rule determined by itself.26:11It's action that's taken that basically email or a notification within the UI and obviously you can extend that whole model to give you a bit more action that within your system.26:31Obviously, without talking root cause analysis, there's NO insight that's meaningful to you, just telling you this is wrong, it's not good enough. Root cause analysis basically use machine learning and.26:46But again using AI, the chat capability to give you summarized statements on what has happened according to the data, give you the full gist of.27:04Why does it have this event? And also trying to see what is wrong. If you have used a dynamics for long enough when you're trying to look for an event, we have this capability of.27:21This might be a problem when you open that screen. This does a similar thing, but with more understanding through the pre existing analysis through the correlation and if you have any action executed per your health rule.27:36Then you will be, it will show it there as well. It will, you can always click the dashboard during anomaly to set the particular time frame of the event happening. You can see exactly what's happening on the flow map on the.27:56If it's a business transaction, you can see the calls per minute error rate, things like that on the particular page. If there is a snapshot that's taken, it will get you to there as well. So everything is linked together.28:13And, this insight will be at least partially enough for you to bring into a war room or trying to have this share with the teams. E.g., if it says it's this particular database, you can pass it to that person.28:30Without going down and trying to drill down the actual problem looking at the code, things like that. This is where you get more value out of the box instead of setting the exact health rule, understanding what's going on and drill down things like that.28:47I'm Conscious on time, so I will run this a bit quicker, so I will run this on my own. In terms of action, it's actually very straightforward when it comes to action cause a lot of times our customers have the insight that will be good enough, they have their own procedure, but what.29:05What we're talking about is the integration part. If you are in any software vendors.29:24Presentation you know integration. It's how you get more value out of it in the appdynamics term, we have many type of integration within Cisco, outside Cisco, but mainly I'd say it will be the telemetry and logs.29:31I won't be talking about security, like the Cisco secure app or the network visibility thousand eyes. These are different topics for different architecture. Although it's available, we'll be talking about telemetry, like how does it pass on information and locks, interestingly, Pradeep has talked about.29:50That, we'll be talking about how we integrate with Splunk. Obviously analytics is available, but we want to.30:11Spend our footprint along as an integration. Obviously dashboard will be for a lot of customer the end point of this whole process to understand everything on a very visualized presentation, and then you have a report coming in on the cadence that you're setting up.30:21So for integration, in America we have run a few of these sessions on really digging deep on how to set up web hooks and to extend that kind of messaging towards to other application. This is towards Slack. Obviously this.30:40Particular screenshot is quite old, but you can have this in the messaging app.31:00Here is an example Slack payload, we have been integrating with them a lot closer obviously due to the acquisition of Splunk. If you're interested in building this properly with a payload, set up a template set up properly.31:08In the community of Splunk, there is a proper article about this, but yeah, it can be very helpful if that's your regular tool within the team.31:29Pager duty is also another very common one that if you have a very proper ticketing escalation system from there, we can just pass this on. Obviously Atlassian is also it's even better, it's fully integrated, you don't even need to have any preexis.31:40Thing payload. You can just set it up on the UI.31:59Here is pager duty payload example. In pager duty template you can see two custom variables you have the token to off with and also the event type.32:04It's not rocket science, but I won't delve too deep into that because there is a dedicated sessions that will happen in America. If you guys are interested, I'm happy to have that run in APJC as well. And obviously Splunk.32:21Lock ingestion through lock observe connect is the reason why I keep saying that is because it's a very recent development. We can use officially supporting the universal forwarder through.32:39Through basically the UI to connect, we're basically having a tag between these metrics and chase them up in both sides. In the app this button will be available in the dashboard application dashboard, just the front page of application. There's a button for that. This is transaction when you click into that.32:56You can link to each other.33:16In the note page, it also have the same thing so it matches each other through a tag again and here same thing transaction snapshot is a great one when you're troubleshooting you want a lock enrichment to look at say e.g. spend, that kind of thing.33:19This is very good. The unofficially it supports more type of connector, so if you're interested in, ok, I want to have lot of connect, but I can't use universal for it for some particular reason like in the AWS.33:37Yes.33:57Kubernetes environment, you might not have the footprint. Let me know. I have some unofficial document that I've set up with with the customer. It can be done. The connection is logical. It, if you have Splunk footprint, this is very helpful and it's a heavy.33:58Every development is in a very heavy development phase, we're trying to have.34:18A more tightly connected system in terms of observability and you'll get great value out of it just by having Splunk lock enrichment. And again, obviously if you have analytics licenses, then you can use FD like analytics. It's still very.34:23Very very useful and you don't need to integrate, you get more information, more insight, but obviously there's license limitation and things like that.34:43Okay, so here are some of the suggestion or summary that I have. It's quite raw from basically in field experience from me to you guys.34:54Number one is garbage and garbage out. This is obviously for the AI and you will know that this is a very common term. It also happened with observability softwares in general. If you don't give the right.35:09Input or the right massage on the metrics, you will never get good insight or good outcome or good action out of it. It's just a mess of data lake, things like that. So make sure you massaged it, you have the right business transaction and things like that set up.35:25Small but nice, not big but bad.35:45So 2nd one is small wins as we discussed. Stack up good business transaction response model and dashboard. Stack them up that way. That will become a better, you'll have a better outcome. Understand.35:49Not understand but trying to make sure that everything you do is within your KPI that aligns with your team's KPI and your org's KPI. Don't waste your time on trying to capture everything and slowly expand the scope. That's very, very important.36:05Last part is there are a lot of customers that doesn't know that there is a huge community outside that can help you. Cisco community is there. We as customer success specialists are there. Sometimes we don't get tagged into.36:24You as in the customer simply because of contract size, things like that, but it doesn't stop you from asking questions from here. It doesn't stop you from, if you want to send me an email, I can help you. That's not a problem. If you have a partner, partner is happy to find someone to help you out and ultimately if you want someone to hands on or do it.36:39For you as a pilot test or just to do the whole thing, professional service available, feel free to consider every single free and paid options and.36:59I can open the, this session and see if there's any questions done so far, ok?37:12Alright, if there's none, I am happy to show you a few things ok you can do like 05:10 minutes.37:30Of, showcasing in our demo, the famous demo environment that if you have attended our sessions, you'd be very familiar. That sounds good, Kevin, we should have some practical.37:45Understanding of the features that we just talked about. Yeah, so can you see my screen?38:05Yes, it's visible. Thank you Kevin. Perfect. So obviously going to the application and choose the right applications, I was here, so it should supposed to show you this particular page. This is such a perfect environment. I don't want to use this.38:12Right, let's use this one. Now, oops, sorry. Now, this is a better environment obviously the smaller so you can argue that, oh, this is not something we can achieve.38:33But take notes from it, like, these have six notes within the TR, one note from a TR. What you want to do is to split the right hierarchy, put them into the right TR, and even split it into applications.38:48AppDynamics is now capable to do cross application correlation as well. You can see it goes to the next application and speaking about it's when you're trying to clean them up, obviously renaming is a good one. But at the same.39:06Anytime you might want to move around the notes.39:26When it comes, when it becomes too many tiers, you want to move them around, put them in terms of function, put them together. But I'll be careful on doing that because the UI can't do everything. You can create an app server tier as a planning process.39:29And but then eventually you still need to go back to say your terminal to to change the config the controller config XML and put it into the right tier. You still need to do that.39:48The and then obviously restart and things like that. The pro this particular function that to create an app service here is just for you to plan things out, see visually how it looks and the naming. Once you're done, you can go back there and do it.40:05If you are trying to do it just by ok in my head and then do it in the XML it's not very and maybe for me it's just not doing it in my head is not very useful. So that's one thing to make sure that this is very clean. This is one of the.40:24Suggestion I have for the customer we I've talked about. Second thing is you can see these, these are not very good naming although the URL has some really good contacts. We still want to make it into something more.40:43Business orientated and then once you look into it, you can, like for us, if you are actually an SRE, you know what's happening once you click into it, you don't need the name to help you to go to understand the URL. So you want that to be a bit more simpler. This will help you in the long term.41:02The reason why I'm saying that it's because when it comes to alert and response, the template will have this BT and you don't want it to be URL when it comes to sending it upstairs or sending it to the business guys.41:22And that happens with everything. Tiers and notes, business transaction, you want to make it nicely and we talked about the lockdown. Once you lock this down, these are the business transaction that's available in this particular screen, and there will be an overflow page happening here.41:38At the moment we don't have any.41:58Then you can see, once you have set up say 15 business transaction, that's helpful, name right, logically correct, have all the health rules, then you start to move those that's locked outside of your monitored page to the monitor page.42:00This is extremely helpful, basic but helpful, and, that's what we have talked about just now. Alert and response. So as I said, you have this particular overview.42:18When we talk about anomaly detection, you see once we turn it on and you have put the app the BT into it or perhaps it's not BT, perhaps it's a.42:35A database, you can set them up and it will start to train when it's ready, it will say ready. These are unplugged, that's why I'm not available, and it will start to catch anomaly.42:53And in each anomaly you can have the whole page of RCA. Now RCA has also, the one click RCA I should say has now been expanded to even your manual health rule events.43:11So I'll even capture anything here. There's NO guarantee, NO, not really, let's see here alright, so say this one.43:29Yeah, if it loads a long time, that means the machine learning is doing its job. No, this one doesn't have any. So this one, if you see this page, that means it doesn't have the RCA. It does show you enough information. It still gives you a dashboard button.43:51Still show you the events, the execution, timeline and all that, and you can zoom it in and out.44:11Things like that. But in order to get an anomaly detection one, hopefully I can find some. Let's see.44:18Seems like our environment has been changing a few things. Yeah, unfortunately, I can't show you any. I don't seem to see any. Let's see if we can find it here. Nope. Yeah. Unfortunately, I don't have it today, but.44:41If you go into an event and you see that for the anomaly detection one, it will surely have it for the menu health rule ones, it's still expanding its scope. Slowly it will cover most of the events and and will collide with the snapshots and things like that. For those who has.44:57Blunk footprint.45:16In the dashboard, you can see this button, it will go to the logs that for this application. If you go to, a snapshot, say this one has a full snapshot, it will follow this business transaction, follow this time and all that with a, with one click, which is very, very helpful.45:20Business transaction, you go into that same thing. Tiers are notes say this one same thing. It will connect to that particular, it will collect, we will collect.45:42Logs.46:01It will build the SPO for you for you to use basically. But yeah, I think i'm conscious of time. That's everything I have. Thank you everyone. I appreciate your attendance, with there will be a.46:02The feedback form after feel free to fill it out, let me know what you want to see, what session you want to run.46:22And, if you go to Cisco community, there are a lot of sessions that's still coming up, including a smart agent, some Splunk observability cloud session that's coming up, feel free to attend. Anything else from you Sanjeev?46:29Nothing Kevin, thank you so much for providing the session on through the demo of how the visibility and insight is provided by AppDynamics and also how action is suggested. Great session. Thank you so much for that. I see NO questions in the chat.46:47But yeah, we can give a minute more, you know, if there's any question from anyone, please feel free to post them. Yep, then we'll wrap it up from here. Thank you guys. Yes.47:07Alright it doesn't look like we have any questions. Alright, I'll close this then. All right, see you. All right.47:22