Wireless Assurance In-Depth on TechWiseTV
♪ [music] ♪ – [Robb] Extensive research that I just
conducted a few minutes ago with the crew here has revealed that wireless, or Wi-Fi,
consistently ranks high on two different lists: Things we expect all the
time and things we complain about the most. Cisco DNA Center has
finally brought intent-based networking to both wired and wireless seamlessly
treating the entire network as one big virtual switch. And now with DNA Center
rolling out Assurance, what does this mean for wireless
networking? – [Lauren] As it turns out, a lot.
Wireless expert Paul and I will cover remediation, wireless
troubleshooting you will actually look forward to and many other tools that make
all aspects of managing, monitoring and troubleshooting so much
easier. We’re going to show you exactly what assurance means for wireless as part
of an intent-based network. Wireless expert, Kathy,
joins Robb right now. – So DNA Center’s Assurance speaks of a
level of confidence when it comes to the entire network. And Kathy,
it’s so good to have you here withTechWise. So I’m curious about,
we talk about assurance with wireless, specifically,
you’re the product manager in this group. How do we begin to speak of,
“What do we mean by assurance for wireless in this context?” – [Kathy] Yes. Sure. So on one hand,
while I suggest an access to the network. On the other hand actually wireless is the
most demanding, unpredictable and is part of the last co-dependent chains of the
services. So wireless actually tends to be the one that get the most plan even though
the [inaudible], maybe, is different. – Right.
– Yeah. – Yeah, because if you think of a wired
environment, of course, everything is much more predictable, your physical layer is
usually pretty easy to troubleshoot from that perspective,
but wireless can have things that come and go, it seems like really quickly. – Exactly. – You guys have mapped out these five
phases that you’re calling, and correct me if I’m wrong,
Correlated Insights. Can you walk me through kind of what’s the
thinking behind Correlated Insights and why these phases?
What’s the thought there? – Sure. So when we look at the entire
experience of the client, we want to look at it from end to end,
so that’s why we map it out into five phases.
So in order to have a successful experience in the network,
the five phases need to be operating very well to make sure the success.
And if you look at the five phases, the list under, each of them represent
Insight. So what is Insight? Insight is actually a very
polite word for issue. – Issue. – It means that if issue happen, it will
have degradation of its own [inaudible]. So I’ll give you examples. So for example, if you have
a association failures, that will impact your client on-boarding
experience. If you have a slow roaming or you have sticky client,
so that will impact your client connected experience.
And we check that all the way through application performance
experience as well, right? So all five phases, all of them are
critical and it’s co-dependent on each other. – That’s interesting. Right.
Because it’s not about a week if one goes away. They all could be precursors to
something bigger happening, and so this becomes,
and by Correlated Insights is the fact that they work together to hopefully let
us get on top of a result before it affects clients or they begin dropping or
complaining, or whatever it may be. So there were three things that I kind of
thought were very interesting that you guys have done differently that really
play into this notion of Assurance, and as I will say confidence that this
level of assurance brings, since we’re using assurance as somewhat of
a product name within this area that I like the connotation as well.
The first of those three is this notion of we’re now pushing communication doing a
kind of a push versus a pull. – Okay. – Can you explain that SNMP versus…What
is that? It’s in my notes. – Streaming Telemetry. – Streaming. Thank you.
So Streaming Telemetry. That’s it. – Yeah, yeah. Sure. So actually for the
first time in the history we’re streaming telemetry data directly from the
controller in the IOS 8.5. And I can tell you why this
is the game-changer. – Please tell me why. – Right. So comparing the Telemetry data
and SNMP, the main difference is that Telemetry data is the push method while
SNMP is a pull method, right? And we have three benefits.
Number one is that it has a lot low overhead when we are pushing
data directly from the controller. – Okay. – The reason number two, or the benefit
number two, is that we can optimize the data when we send, so we can optimize in
terms of Kbytes or event. And lastly, and to me this is the most important one
and it’s how we change the entire industry, is that we send data a
lot faster, right? So right now we can detect the issue in the order of seconds,
whereas if you compare with the traditional SNMP data,
the issue we detect five minutes ago, so it’s not really helping the users. – Excellent. Okay. So,
we get this information faster so that reactions can to happen faster,
so we get on top of stuff faster. It’s fed by these Correlated Insights as well I would imagine.
– Exactly. – Okay. So that’s Streaming Telemetry.
The other thing was this notion, and this was really big, you see, our
partnership with Apple, which we’ve talked about on the show here before.
And we talk about analytics now, though. We’re getting more information from Apple
devices running on a Cisco network that we haven’t had. What are we doing now? – Yeah, this is something that all of my
customers are excited about. So this is very, very unique about iPhones
and iPads that are running in IOS 11 on Cisco network. So for the first time
that we get access to the data of the real actual user experience.
So the device send us the actual user experience. And this is something new
with Cisco and that’s why we call this Cisco Wi-Fi Analytics for IOS.
So there’s three things that the Apple device send us today that this is new.
So number one is that it send us the BS ID, it send us the SSI.
So entirely everything is about the insight, how the client view
about the network. – How it perceives its connection. Okay. – Exactly. So we have the infrastructure
view of our client, but now we know that how the client view about our network.
So that is something new. – Yeah, we never get that because
standards don’t allow us to do that. We don’t have a level of trust.
We’re not going to trust clients to tell us that.
– Exactly. – But we do with Apple. – Yeah.
– Okay. – And then the second information that we
get is about the hardware models and the software versions that it’s running.
With that information we can do a lot of, like, correlations insight about the
hardware model or the software versions. And lastly, and to me that is also the
most important one is that the reasons why the client disconnect.
So it inform us why he disconnect and we know what is the cause for that.
For example, it could be the services root cause, for example,
that could be triple A, section terminal or that could be
validation failures. Or it might be just the user actions like,
“I turn off my iPhone, or I disabled the Wi-Fi,” so that’s why it
disconnect. So now the infrastructure actually know the reasons
why the user disconnect. – I like that. And so obviously we’re also
getting neighbor list from the client perspective so that now we’re not
dependent on the APs telling us, “This is where we think you should be
roaming and we’re taking into account.” This all comes together through DNA center
now so we can see this coming together. Now, speaking
from an analytics perspective, I can’t help but notice you’ve got…What
is it you have in front of you? – Sure. So this is the one that we
launched together with the DNA Center. – Okay. – So this is the Cisco Active Sensor.
So actually, it’s very light and small. You can see it’s like…
– Yeah, it’s nothing, yeah. – …the palm of my palm.
And then you actually can carry it and then you can change it to AC power so that
you can plug it anywhere. – So if I’m having trouble,
literally… So if I don’t have an Apple device, it’s giving me
perspective from the client and I’m troubleshooting a specific location,
a dead spot, I get the client perspective, plugging this in AC in a wall or at a
computer level. – Exactly. So you plug it at a floor level
so that it’s very close to your real client experience. And actually many of
our APs can become a sensor as well. And it can be done automatically,
you don’t need to do it. The network will do it for you when it
need to do or you can schedule it to do it for you. – That’s right. Certain APs actually will
do flexible radio sensing, I think. Is that correct?
– Yeah. – And then they will convert,
and so that one’s at the ceiling level as opposed to this at the client level. – Exactly. – But all this again comes together in a
kind of a simplified delivery through DNA Center so that we can
hopefully troubleshoot either very quickly or much more quickly than we could before,
or ideally prevent things before users even become aware of them. Correct? – Sure. Yeah, that’s correct. – Final point. What do we need to remember
here? A summarization of what’s most important. – Yeah, sure. So wireless has been
dramatically improved in three level. So level number one is detection.
So right now we can leverage the entire network, sensor to detect the issue
for users, so help them to do faster troubleshooting.
Level number two is prevention, so because of the sensor proactive
modeling we can prevent those issues happen for the users. And lastly,
in term of suggested remediation so we can have to suggest the action for user to fix
the issue if it happen. – Perfect. And I think that’s actually
Lauren’s favorite. So thank you Kathy very much. It’s time to go over to the lab
where Lauren’s going to break this down with us. I believe she has a special
guest with us to see what this looks like. Let’s do that now. – You know, what I love about Assurance is
that it lets us, not only take the wired network, but the wireless network
and gather all those analytics. And here to tell us more
about that is Paul. – [Paul] Thanks Lauren.
So Assurance has so much amazing information on wireless network from an
access point controller and down to the client itself. So I’m glad to be here to
talk to you about that. – Let’s dive in. – Yeah. And here we’re going to look at
the dashboard. And at start of the day, we’re able to see high level and drill
down into all the various sites across the world, right?
And here we can see the different sites by color schemes and be able to see what’s
more problematic, and we can drill down from that. Next,
we’ll be able to show you from those sites, what are they made of,
right? So from a site, we have, for example we take San Jose as we’re
here, and we’re managing the Building 14 which is next door to us. – My old building. – Yeah. It’s ours too.
And what happens is that we have four floors and all these four fours we have
all the access points and here you can see a really nice visual look at the heat map.
So you can get a good idea of the coverage of all the clients. – Yeah, I love this.
So if I’m maybe over here, maybe I’m not getting the right coverage
or something like that, and you can see that pretty easily there. – Exactly. We can hover over this and see
who’s it talking to, what are the neighbors are seeing and how
much coverage they get. – I think wireless admins
are going to love that. – I know. And from there they can actually
look at specifically, what are the problem access points. So if we’re looking at the
wireless network and we can be able to drill down and look at, for example,
some of these access point might have issues, so we would be able to have
Assurance bring up all of the problem devices, these access points, for
example, and we’ll be able to look at this particular access point
over a period of time. – That’s awesome. So even if I had a
problem yesterday at 3 p.m., you’d be able to go to look at that,
see what the problem was. – Right. And you can see there’s ups and
downs, right? And it gives you an idea when those are and we can actually drill
down to them and look at, in a historical sense,
what are these trends. – And yeah, the trending analysis being
really important there too, so you can do it for the last seven days,
every day at 2 p.m. – Exactly. – And that something’s happening, right? – Exactly. Right. I mean, we can have,
for example, over lunch time somebody might be firing up their
microwave ovens, things like that. – Good point. – Right? And then cause a lot of
interference for wireless and we would be able to have Assurance provide very good
insights for that. – Yeah. Robb’s downloading his Netflix
movie, it’s taking everyone’s bandwidth. – Absolutely. Absolutely. So with
that, we can drill down to from the network health, we can go right into the
client health itself, right? So here we can look at this
particular building while we’re getting a lot of clients. I mean,
Assurance is able to do this across all the different sites and be able to
correlate them and give us an idea of what are the problem devices. – That’s awesome because I imagine,
you know, it’s telling you the problem devices, but whether you’re running some
sort of enterprise or maybe it’s a conference-level thing where there’s a lot
of high-density networks going on, that can be really helpful. – Absolutely. High density is a great
example because, for example, we have, you know, how long it would take if we’re
such…so many clients are in one space, how long onboarding times would be?
A lot of them might be connected, but, you know, at the same time they may not
get an IP address. And you and I know that a lot of people don’t know whether or not
IP address may… – It’s the wireless. – Exactly. It’s the wireless. It’s always
the wireless that’s a problem, right? So this is a really useful tool for that.
And same with the network devices, we have clients, problem clients get
bubbled up and gives us a very quick view into what’s wrong with
each of the devices. – So yeah. For a network device, it’s an
access point or something that’s a higher priority, too, but if someone just doesn’t
remember their password, well, maybe that’s not as important, right? – That’s right. Right.
Chuck Robins you may be checking… – Maybe that. – So with that, let’s find out for example
we have a client… a user call us, right? We search for a particular user that calls
us. “I have a problem. I can’t get on Jabber,
I can’t get on Wi-Fi. I’m in this particular area.”
So we’ll search for this user. And this is nice because we also provide
not only a Device360, but a Client360, so we can actually get an in-depth view
into what the client issues are. – That’s awesome. – So here we’re taking a look at Rashid
who’s carrying multiple devices here because he’s a very productive, you know? – Right.
– And for example… – The more devices, the more
productive it is? – Absolutely. That’s how we…Right.
And from here we can look at over time again, this is sort of the
unique theme of all the historical view into the client performance and he has
multiple different devices consolidate in a single view. We can look at multiple
different ones like iPhone, Windows machine. We talked
about IOS devices, right. – I noticed a few Apple
devices or IOS devices. – Yeah. – Yeah. What can we do around that?
Because I know there’s a partnership between Apple and Cisco. – Right. And this is very special, too,
and unique to Cisco, is that we have Apple Analytics and that gives us, for example,
you can imagine, we have a way to have the device talk to the network.
So we have two-way perspective, very in-depth specific issues that are
reporting from the device itself. So let’s take a look at what that would
look like. So here, Lauren, we can see that, for example,
this particular device is an iPhone, but it’s not only a iPhone, it’s an
iPhone 7. So it could be iPhone 8, iPhone 10. From there, we also have the
operating system. And what’s good about that is because over time we may figure
out that perhaps, you know, OS update, a firmware update may affect the
network in one way or another, so having this is very useful. – Right. It all comes back to analytics.
If everyone else is running 11.2.1 or whatever, and that’s the
problem child, then we know. – Yes.
– That’s great. – And take a look at the IOS Analytics,
other IOS Analytics. Here we can see those events that we
talked about. For example, you know, it could be a triple A,
radius user authentication credentials or perhaps it’s just someone turned off Wi-Fi
for airplane mode, those type of things. So we’re able to get that
reported from the IOS device itself, and on top of that we get the neighbor
list. So it has an intelligence to report what are the nearby APs and be able to
have that list that’s optimized so that they can move around and roam. – So what are you using to collect all of
this information, then? – We’re using just the network
infrastructure that you already own. – And other sensors or what else is
happening there? – That’s a good point. So, Lauren,
you talked about the sensor. So we’re going to go into Proactive
Insights, and this is an example of the 1800 sensor that we have.
And this is so small. You can pretty much plug it anywhere on
a table and have that floor view as if it’s a client perspective, right?
Moving to sensor management, we can do that. We can use these sensors
or we can use existing APs in your network by intelligently turning them into a
sensor or one of the radios that are not being utilized, and that can also serve as
a sensor while serving data itself. – That flexible radio. – Yeah.
– MRA, right? – And it’s very unique to Cisco as well.
We’re great to be leading that. And so, for example, these tests that we
set up for sensor, we can have various different performance indicator tests
synthetically test, for example, how they’re getting on the wireless network,
how they’re able to onboard themself and how long it takes, as well as doing radius
tests; reachability, mail test, as you can see here, web server test.
A lot of different tests. And we can explore that here. But for example, we can
see why is this failing and also look at specific time sequence of what are
these failures mean. So for example here you have DHCP IP
address is slow, and that would be able to be given to us by the Proactive
Insights with sensor. – So when someone blames the wireless,
we can say, “Nope, it’s the server guy, that the DHCP server is wrong.”
– Right, right. – That’s awesome.
Thank you so much, Paul. That was fantastic.
– Thank you, Lauren. Yeah. – Assurance for wireless only on Cisco’s
DNA Center. Big thanks to everyone for watching. ♪ [music] ♪