What Does Network Monitoring Mean in the Cloud?

What Does Network Monitoring Mean in the Cloud?

Show Video

[Music] thanks for joining us to talk about what does network monitoring mean in the cloud my name is brandon shopp and i'm a gvp of product here at solarwinds i've been at solarwinds uh for about a decade now and my background is in network engineering so i started out as a network engineer before i joined solarwinds and and i actually shifted into software and had been the product manager for the orion platform network performance monitor and many other products here and the solarwinds family with me today i have satish satish go ahead and introduce yourself hi everyone uh welcome uh to our session here uh my name is satish grandy i head the office of the cto at solarwinds my experience is has been in the networking field for about 15 to 20 years i started my career off as a network performance management uh developer at neutral networks went into protocol development across the entire networking stack with a lot of experience on the in juniper on the sdn and the nfe space and uh did a startup after that with being the first hybrid cloud management platform that was dealing with cloud and the migration to the cloud and of the applications and then i was the chief architect at pulse secure looking at zero trust technologies and the sassy and sd-wan technologies um i believe uh we're gonna kick it off with market trends brandon um on to you yeah that sounds great let's uh go ahead and dive in and talk about as you said some of the market trends we've been seeing out there and so you know in talking with customers one of the things that we are hearing and seeing more about is the the leveraging of clouds and as well the you know moving of workloads uh from on-premises traditionally uh over to the cloud and so you know with these trends along with several other trends that are driving the evolution of the the network performance monitoring market you know network visibility is critical uh you know even in a shared responsibility model where the cloud service provider is ultimately responsible for many aspects of the network infrastructure you know as an end user you are still responsible for securing configuring and kind of maintaining you know those services and how you're leveraging them uh yourself and so uh we're no longer constrained you know to our infrastructure and the services residing solely you know kind of within the virtual four walls of our corporate environment you know hybrid cloud architectures services and the apps are driving the need you know for npm tools to support those end-to-end views of those transactions not only in a traditional on-premises uh environment but also you know in hybrid and in pure cloud environments you know additionally to that you know there's been broad adoption of you know software-defined networking technologies and nfv technologies uh you know such as sassy and sd-wan and you know others are are ultimately automating or driving organizations uh to automate the provisioning and you know forcing uh prevented performance management tools to also uh follow as well so satish you know what are you seeing out there from a trend perspective yeah brandon as organizations are evolving and introducing these newer technologies into their networking stack and operation stack what what this has done is has increased the complexity of their environment right be it the move to the cloud and it's not one single cloud that the end customers are moving to they are actually you are doing multiple clouds right and and they cannot lift and shift their existing environment so they are also having a data center technologies and private cloud technologies uh they're also moving some of their business critical applications such as microsoft outlook exchange to office 365 so the sas applications are being adopted more and what the pandemic has done is changed the way we're working so a lot of their employees are moving to or from the office environments to their home environments and there's a lot of offloading that is going on from remote offices or home offices directly onto the applications and applications are also evolving to take advantage of these cloud native technologies and that's also introducing a lot of complexity and what um this has also done is there's a lot of control that the that used to have the it operations team used to have they're losing control of some of these environments because of the cloud providers providing some of these technologies and the sdn technologies layering overlay technologies on top of it so so what this is also causing is a humongous amount of data being generated and a lot of automation that needs to happen to solve this problems right so uh machine learning based uh algorithms to assist the humans is also uh going to play a huge role and that's where some of the innovations are happening right so but first before we get into the new use cases that we that these cloud technologies and the newer technologies uh brandon could you give us your perspective on how the on npm tools have evolved over time and the on-prem uh use cases that they used to solve and then we'll get into the cloud use cases that are being introduced yeah no absolutely and so historically you know network monitoring and even other discrete you know management tools allow the network admins to collect various metrics off of their infrastructure you know within their their virtual four walls you know the routers the switches the firewalls you know i know when i was a network admin you know 15 20 years ago you know i was kind of really focused on you know my section of the world you know the the actual network infrastructure and and just making sure that it was performance and healthy you know is it up is it down so you hear you know if you look at the f caps model you know uh performance and faults it was really focused you know around those elements there and so you know nowadays though uh it's not enough to ensure you know that you know things are up and that's um they're online and that's there are no obvious you know performance issues within the infrastructure you know you really want to focus now so on the end user experience and so we'll talk more about that um you know and you know it's not enough just to ensure you know that proper user experience but you also need to make sure that the the network and the underlying infrastructure is that you have the proper visibility you know into that in order to deliver those uh those services on that front and so you know one of the things that we've started to see from an on-prem npm perspective is you know we talked uh you know a little bit about uh apis and and leveraging more of apis by vendors and so you're seeing that in some of the newer technologies historically you know you know the networking world has been a a very snmp driven world um you know if you just write to the you know the ifmib uh as a monitoring vendor as as new hardware or new technology comes to market then it just work it it just works it's just supported but nowadays you know apis provide a lot more richness and flexibility uh you know to the vendors themselves and so they're exposing more and more data and endpoints uh you know of information through an api and so you know with that you can stream that and you're not just no longer just looking at a a certain point in time you know every 10 minutes i pull and give me the statistics off of this device every eight or ten minutes now i want it to near real time uh you know because you know minutes can mean millions of dollars to my business and downtime or it could mean the difference between uh a consumer purchasing something from me versus that consumer going to a competitor of mine because of a poor user experience and interacting you know with the technology uh with my business and so um so that's you know some of the things that i've been seeing on the on-prem side but let's talk a little bit more about you know how cloud is driving use cases today you know within it organizations so as organizations have uh adopted these clouds and moved some of their applications to the cloud and also also adapted sas services predominantly and with the pandemic playing a huge role and a lot of on-premises vpn concentrators are getting overloaded with traffic some of the sd-wan technologies have evolved to a case where the customers are able to offload their uh employees access to these sas services as well as the uh cloud applications from their remote office edges right so this will also introduce complexity into how they have their end user experiences evolving so with our uh networks and uh that's where the relevance of network synthetics are uh taking a um come back right of sorts right and they've been pushed to the forefront right uh to uh track of the you know hop by hop analysis right um and that that's some of the newer use cases that are propping up capturing the end user experience of these remote employees and work from home employees similarly with all this data and all the dynamicism that the cloud introduces there is a lot of use cases that are evolving for real-time monitoring and via streaming telemetry collection mechanisms that the cloud service providers are putting in play on the streaming telemetry as we'll talk about uh in the key new data sources that are being introduced um are bringing in newer use cases for um real-time monitoring and with all this complexity brings in a lot of data and a lot of our dependencies that need to be tracked correlated across the infrastructures and the applications and that's where um the role of machine assisted learning right for for the humans to make uh effective root cause analysis and uh the field of aiops as a new thing is emerging right so as uh alluded to there are a lot of new data sources that are cropping up in addition to the traditional data sources that we have brandon uh what are the new what are these new data sources that that are coming into play in this new world that is evolving yeah no that's a great question and before i jump into the new data sources the one thing i want to make sure that you know you and i have been talking a lot about apis and streaming telemetry and and some of those newer protocols are or or data sources that we're going to dive into but that doesn't mean that snmp for example is going away uh you know snmp is still going to be very much a trusted method in which to go get uh data for certain use cases and certain needs and so same thing with flow-based technologies you know netflow and jflow and s-flow et cetera those aren't going away if anything you know that's evolving that's you know uh augmenting if you will you know you talked a little bit about uh like an aws with vpc flow logs you know and so being able to stream those in and and visualize that along traditional netflow based traffic and and so you're still going to have you know those traditional data sources but as we've talked about you know there are going to be additional data sources and really focused around kind of that that end user experience uh angle you know so looking at down at the packet level uh and being able to understand you know from a packet perspective you know uh both at a troubleshooting level as well as at an observability level you know uh you know what traffic is going down what route uh you know but also you know as we look at uh streaming telemetry as you're talking about some of the hyper scalers you know aws azure gcp you know you want to have real-time visibility and so you know the one thing that i look at is you know the days 10 15 20 years ago when i was a network admin and i was focused on just making sure that the infrastructure was was performance and and was running uh nowadays you know i need to focus and make sure that you know my end users experiences are are positive ultimately and and again going back to some of our previous points you know a polling based methodology doesn't give you always the clearest and full picture uh there at that point and so that's really where streaming telemetry can come into play is giving you you know more granular near real-time data stream you know for troubleshooting issues as they occur you know as organizations try to shift further from being you know reactive based to being more proactive you know you need to have that real time or near real-time view into your infrastructure into your applications and ultimately into the data underneath them uh so you can you know catch those issues before they impact your users and address them uh and so but all these you know we've talked about user experience a lot and and you know again user experience is not a a new concept uh it's been out there for many years but more so at the application level you know looking at real user monitoring with rum uh or synthetic monitoring uh from a you know from a bot perspective and so but there's an element of network uh digital experience and so you know what's the end users experience at the network level and where are they potentially getting seeing latency uh or seeing uh other issues you know within the transport mechanism you know uh in interacting with that application and so you know having uh technology and capabilities that can look at that that that network uh traffic or network synthetic level you know using technology such as our netpath technology uh which you know gives you visibility not just in your infrastructure but also into your isp's infrastructure and even into the end provider so let's say you know again we've talked about 365 you know from microsoft here a couple times uh already and so you know if i want to understand from my various sites uh you know you know how is traffic uh getting to 365 and and where our potential bottlenecks are issues and so again that way you can be proactive you know who to call you know where the problem's at is it my problem or is it microsoft's problem or is it you know my my uh service provider's problem so the quicker you can root cause that uh the quicker you know you can get that ultimately addressed and solved you know for your end users and so um you know the other thing too is you know the move to sas and and and the various other trends you know in moving applications to the cloud or the adoption of newer wan-based technologies you know is you know not giving you kind of the the control and the granularity that you want you know for monitoring and so that's again really where this synthetic looking at it from not just you know lower in the osi stack but looking at it from a layer seven perspective as well you know you can kind of get that complete and full picture on that front and so you know with the the you know the dependence and the need you know the uh you know the criticality of the infrastructure and the applications you know to your business and to your end users you know the other thing that you know we've been seeing as well is that you're talking to different people within the business uh you know different people can be impacted so it's no longer just you know the traditional you know network admin or or it operations you know uh type person you know within an organization there are other people now that are impacted ultimately and so satish you know you know can you shed some light into that around kind of what types of folks ultimately um you know are going to be impacted here right so brandon we touched upon in the market trends there's a lot of market trends are going on but we didn't touch upon how the enterprises are evolving uh with uh in dealing with these market trends right so as the adoption is increasing there are net new personas that are appearing in this whole thing right where be it the cloud architects right there be it uh centers of excellence that are being established across organizations with the cloud center of excellence where the people from i.t operations network operations security and cloud operations are all coming together in going uh helping the enterprises through this digital transformation in the end it's a line of business owners whose applications are being moved to these newer architectures and being facilitated by the net new architectures at the infrastructure and network level so the line of business owners are also getting affected and that's where their end user experience for their customers orbit the employees of internal employees that are using these applications that are um so the line of business owners are becoming a key personas that are getting engaged with the netop similarly security secure security is a huge thing right um a lot of enterprises are getting affected from a security aspect with external threats coming in so the network performance monitoring with all the uh data that it is collecting is can collaborate with the security people whether be it from a east-west movement of uh these ransomware kind of things or viruses right all the data that gets collected at a packet level flow level can be provided to the security teams for further analysis right or which end points are connecting to the network really which uh how the data is coming in going out right will help the security ops people uh with with their thing with their um with trying to get the uh borders locked down or the perimeters locked down or identifying uh threat actors that are present within the network so these are all the people and not uh when we talk about all this transformation that is going on that one thing that um everybody is going through is the agile transformation with the infrastructure as a code the whole devops secops uh revolution that is going on so those those are the ones when when people are moving to the cloud the devops and the sres are all the newer personas that are coming into play so these are all the net new personas that uh have are either the existing personas are evolving to or net new personas that are interested from the data that is being collected from the network performance monitoring tools right so um we've talked about a lot of things brandon right we talked about market trends and uh where we talked about what are the net new use cases how historically things have evolved what are your key takeaways uh and the key uh things that you want our audience to walk away with from from this uh things that we've talked about yeah no it's a great another great question and so you know when i talk when we talk about you know the impact of of cloud on network monitoring there's a few things that i i ultimately take away number one the amount of data is just continuing to grow uh that or you know that uh technologies you know are are having to absorb and and you know mining through that is becoming a bigger and bigger challenge for organizations you know two the the means by which you know technologies are exposing that information to us is continuing to evolve as well you know so we talked about uh you know streaming telemetry and and uh some other methodologies which again maybe haven't uh aren't necessarily new but they're they're finally starting to gain traction from an actual um you know deployment and production usage but you know i think the other thing too here that that you know that sticks with me is that you know the hype cycle or the technology you know cycle is compressing it's moving faster um and so technologies new technologies are coming to market all the time and so uh you know with that again that's just continuing to compound and add to the data explosion and so as you look at that you know ultimately you need something that's going to help you kind of sift through that data otherwise it's going to be a needle in a haystack i mean we thought that uh you know network monitoring was a challenge in the past you know now it's much more you know it's multiplied uh in many levels and so you need some technology you know that is going to help you really sift through that no you know that noise or cut through the noise uh and that's really where ai ops can come into play uh you know that we talked about you know with ai and machine learning you know what are the things that i truly should pay attention to and so you know that's another thing that i'm taking away here as well is that you know uh you know machine learning and ai ops is is absolutely a technology that you need in a you know modern observability stack today and so and luckily enough you know that's something that that solarwinds is actively working on on that front so but i'm curious you know uh you know what are your primary takeaways from from today uh uh satish yep um great uh some um uh takeaways from from this brandon um so some of the takeaways that i'm taking is the uh with all these uh changes in uh and the technologies and the market trends that are happening the complexity all right or that the network and the operations people have to deal with is ever increasing so a tool uh that provides them with a single pane of glass that across their full stack right uh that helps them to uh start from their end user experience to the infrastructure level right and correlate this data as it is being generated across right uh across the clouds across the full stack um is a key right and uh the need for real-time monitoring right and the streaming telemetry uh data that is being generated in the dynamic nature of these environments is another key takeaway right so and then the automation of this end end to end right um because of all the complexity involved it's very hard for a single human to go through and sift through all this data so as you touched upon ai ops is going to be a key role play a very key role uh starting from doing anomaly detections or to predictive analysis right or predicting failures right uh so so that it's not the end users that are um seeing the uh this for the first time but it's uh the operators that are seeing it and resolving them proactively right uh so and similarly to predictive capacity planning right um solarwinds has had a capacity planning for a long time and doing this at a pr and predicting this capacity in the future for all the net new things that can come into play with the whole itot convergence uh that is happening there's a lot more data that is going to generate it with all the internet of things that is going to come in so it's going to be a very very exciting decade for the network performance monitoring tools to come forward right so um that's that those are my key takeaways brandon and um that's uh i hope we've gone through a lot of things and i hope our audiences are going with their own key takeaways we would love to hear from them and who i hope they have enjoyed and thanks for tuning in for this presentation absolutely appreciate everybody's time and thanks for joining us and look forward to you know interacting with you all in the future on whack

2022-03-14 02:00

Show Video

Other news