Hi, My
name is Mav Turner and I'm a Sales Engineer at SolarWinds. In this video, I'll be demonstrating the SolarWinds
IP SLA Manager module. IP SLA is a
feature of Cisco IOS devices that allows you to gain insight into your network
service level agreements by generating traffic at set intervals and giving you
details about how long that traffic took to traverse your network. The SolarWinds IP SLA manager module allows
you to manage and create those operations.
For more information on IP SLA, feel free to visit SolarWinds.com.
Now
let's take a look at what an IP SLA operation really is. For example, a TCP connect operation is looking
at a source device. In this case, a Nexus 7000 and a destination device of SMTP
gate .log .tex and seeing how long it takes to get on from port 21 from the
Nexus to this SMTP server. The last time
we ran this operation it took 100 milliseconds to get from the Nexus to the SMTP
server. If we look at the graph on the
right we see that this value has varied greatly during the day. It's gone from 100 or 200 milliseconds all
the way up to 1800 milliseconds. So, if
you have users that complain email was slow during lunchtime at 12 pm, you can
also correlate that to an increase SMTP time response.
IP SLA
operations can be created for many types of traffic. You can also use and monitor hosted web services by looking at HTTP
operations. In this case Salesforce.com,
which looks like it’s actually exceeding its critical threshold. Specifically in the transaction round trip
time; with the HTTP operation we see that we have additional gauges, we see the
DNS round trip time, the TCP connect time and the transaction round trip
time. The DNS round trip time is how
long it took for DNS resolution to take place.
The TCP connect time is how long it took for the server to connect to
the other device. And the transaction
round trip time is how long the server took to process the network
request. If we look at the graph on the
right again, we can see more detailed information about how that operation's
time has varied throughout the day.
Back on
the IP SLA manager homepage, we can gain an overview of all the operations in
our network from the top 10 website. The
top 10 website gives us a great view in how all of the operations in the
network are doing. We can quickly see all
the top 10 operations that have DNS resolution issues due to thwack.com and we
also have an HTTP operation issue. We've
just looked at the HTTP operation, now let's take a look at the DNS operation. It
looks like its taking about 560 milliseconds to resolve thwack.com.
Coming
back to the IP SLA manager homepage, we can also see the VoIP tab and the web
tab. The VoIP tab is extremely useful. You can break down individual sites and call paths
to gain detailed statistics about the call quality users will be
experiencing. We can gain insight into
this by looking at the jitter, latency and packet loss and using those values
along with the specific codec G711 or G729 to look at the MOS, to get a general
idea of how this user's call experience will be. It looks like this issue is specifically
related to latency. We can see the
hundred seventy millisecond on latency, the jitter five milliseconds isn't too
remarkable. The packet loss at 12 percent is not so great either. All of those combine to give us a MOS of 2.9,
which is not good for voice users. What
this allows us to do is to graph these values over time so that if you get a
call that an hour ago, the voice call quality was poor you can come in here and
get detailed statistics about what specifically caused that poor call quality
rather than having to ask the user to tell you next time immediately when
there's a problem.
Now
let's get a little bit deeper and see how these things work. I'm going to click on the IP SLA manager
settings. In the settings tab, you will
see that we have the option to add new operations or edit existing ones. We can also manage IP SLA nodes. Managing nodes is important because after you
add a node into the main NPM product, you have to specify that it is an IP SLA
node. After you specify it as an IP SLA
node, you can then use the IP SLA manager to add and manage the operations on
that device. We can do this by
automatically discovering all IP SLA capable devices in NPM or by manually
adding nodes to IP SLA; but let's focus on the operations.
I'm
going to edit an existing operation. Use
the same operations we saw on the web page earlier. Here's our DNS resolution for thwack. I can edit this to show you the advanced
details. Of course, we have the name of
the operation and can add an additional description if necessary. The frequency tells us how frequently the IOS
device will attempt to resolve this address.
By default we'll do this every 5 minutes. We are
going to make sure that the DNS server of course we're going to make sure that
our testing gets a specific DNS server so we can specify that as well as the
host we're looking to resolve. The
thresholds are configurable on a per operation basis. So, if in my opinion this should be resolved
in a hundred milliseconds, it’s easy to quickly make that change. We do have advanced options if you're using
VRF or any sort of virtual routing. You
can actually specify an instance name to help you make sure this traffic is
going how you want. We can also specify
a VRF to use a specific routing instance if you're using virtual routing.
Let's
look at the VoIP VDP jitter, one of the most popular ones particularly around
voice environments. We see as always we
can give it an operation name, a description, and the frequency that it runs
at. We also see here that we can specify
the network port that we'll be using to generate this traffic. And of course our thresholds for the various
components MOS, jitter, latency, and packet loss are all configurable. Here in the void VDP jitter operation we're
given additional options allowed to set specific codecs and this helps us
calculate our MOS and also test the network for exactly where our end users
will be seeing. As well as a defining
types of service and of course again VRF values. Now again if we want to use a specific
routing we can use the source address that we like this traffic to be generated
from. Now let's add a new operation with
the IP SLA manager module, adding operations is quite easy. You simply specify which operation type you'd
like to see, for example let's do VoIP UDP jitter, you specify the topology
that this will take place in. Most users
are in mesh MPLS environment so selecting a fully meshed path will be
sufficient however if you do need to specify a custom path, you can always specify
the specific nodes that the network traffic should be going through. Then we'll look at all the IP SLA capable
nodes this is what I was referencing earlier in saying that we need to make
sure we mark a node as IP SLA capable before we can create the operation on
it. But once it's in here we simply
select it and click next, now again we'll see the advanced detail options here,
we can set the frequency that this operation will occur at, specify the port
number, as well as the thresholds for the MOS, jitter, latency, and packet
loss. And of course the codec, type of
service and the VRF. So we'll get a last
screen that will show us exactly what we're creating here. We're creating an operation from Austin 3 to
the 2951 and we'll also be creating an operation on the 2951 back to Austin 3
so we can get both sides of this operation.
Now there's something more I want to show you in the IP SLA manager
settings tab, the ability to monitor existing IP SLA operations, so at this
point I'd simply create an operation and the operation would be created on the
devices. Now when we create the
operation we're actually using an SNMP set against a device so we're not telling
in the device and inputting in specific configuration lines or doing anything
so the other manual configuration, we're doing SNMP sets against specific
values to create this operation so you do need SNMP write credentials on the
box. Let's cancel out of this
operation. I'll show you one more thing
about adding a new operation. If you
manually create you operations, you can simply select to monitor those existing
operations. After you select to monitor
them, all you have to do is continue to go back through here, in this case we
use a TCP connect as an example, select that path that you'll be using, and
then the device that the operation exist on.
Now when you select to monitor an existing operation, you will need to
know the operation number. Again, all of
the defaults that we will create here, as I mentioned you would need the
existing operation number in order to completely add this operation within the IP
SLA manager module. Now it is beneficial
to create the operation with the SolarWinds IP SLA manager module because when
you remove it from the SolarWinds IP SLA manager module we could have it be
removed from the network at the same time.
Let's go back to the home page.
As always with SolarWinds products you can customize this page by adding
or removing resources. One particular
resource that's important in the IP SLA manager is the network map. With this we can add IP SLA operations and
show you graphically the status of those operations and how long network
traffic is taking so we can drill immediately to the network issues. We see again the transaction round trip time
is exceeding its threshold. Let's dive a
little
bit
deeper in the IP SLA settings tab. With IP
SLA manager we can also add new CISCO Call Manager to get information about the
VOIP infrastructure, registration, rejection of phones and gateways and other
useful information. Coming back to the IP
SLA manager tab here we can see the VOIP infrastructure and scroll down and see
the VOIP call manager so we see we have Austin subscriber 1 and publisher 1 and
the phone rejection and unregistrations as well as the gateways. We can also expand our VOIP infrastructure
and get total view of what's going on for our CISCO voice environment. Heading back to the IP SLA settings, we have
some system wide settings that can also be modified. We can set the VOIP UDP jitter at a global
level, we can set the default codec as well, and we can also specify how
frequently we'll poll the network devices for the data that they've been
collecting. Of course we can configure the data retention, you can configure
the MOS advance factor and the type default type of service, and of course you
can restore these defaults if you've made changes and want to go back to the
default settings; And last but not least we can look at the database details,
in the database details page we can see basic information about the database
that IP SLA is running on.
You
will find more detailed information on IP SLA Manager at SolarWinds.com. As always, Thwack.com is a great resource for
any SolarWinds products. Thanks for
watching.