WEBVTT

0:00:02.820000 --> 0:00:07.800000
 Hello and welcome to this video
 titled SNMP components.

0:00:07.800000 --> 0:00:11.620000
 In this video I'm going to talk about
 the components and architecture

0:00:11.620000 --> 0:00:16.460000
 that make up SNMP and we'll also briefly
 look at some of the main popular

0:00:16.460000 --> 0:00:18.800000
 SNMP message types.

0:00:18.800000 --> 0:00:23.060000
 So let's start by talking
 about SNMP components.

0:00:23.060000 --> 0:00:26.820000
 So whenever you learn of any new protocol,
 regardless of what layer of

0:00:26.820000 --> 0:00:31.060000
 the OSI model it lives at, the first
 thing you have to know are what are

0:00:31.060000 --> 0:00:35.060000
 the devices that implement this protocol
 and do these devices play a special

0:00:35.060000 --> 0:00:38.840000
 role? Is there something that they
 do that other devices working with

0:00:38.840000 --> 0:00:41.360000
 the exact same protocol don't do?

0:00:41.360000 --> 0:00:44.000000
 And then clearly there's probably a
 bunch of acronyms and terminology

0:00:44.000000 --> 0:00:47.160000
 you have to get used to as
 well with that protocol.

0:00:47.160000 --> 0:00:49.060000
 Same thing is here as well.

0:00:49.060000 --> 0:00:54.040000
 So with SNMP it's really a
 client-server architecture.

0:00:54.040000 --> 0:00:56.720000
 Just like so many network
 protocols are.

0:00:56.720000 --> 0:01:01.120000
 Now in the terms of SNMP things are
 a little bit reversed though in that

0:01:01.120000 --> 0:01:05.420000
 you will have a server which
 we call the SNMP manager.

0:01:05.420000 --> 0:01:07.200000
 Now that's going to
 be an SNMP server.

0:01:07.200000 --> 0:01:11.280000
 It could be a Linux or a Windows server
 and it's running some sort of

0:01:11.280000 --> 0:01:14.120000
 network management station
 software on it.

0:01:14.120000 --> 0:01:19.020000
 And there's lots of software out there,
 SolarWinds sells some, there's

0:01:19.020000 --> 0:01:20.840000
 something called PRTG.

0:01:20.840000 --> 0:01:21.720000
 You can just Google it.

0:01:21.720000 --> 0:01:26.660000
 There's lots of SNMP NMS
 software available.

0:01:26.660000 --> 0:01:29.120000
 Some of it's free, some
 of it costs some money.

0:01:29.120000 --> 0:01:31.320000
 Now that stuff is running
 on a server.

0:01:31.320000 --> 0:01:34.620000
 So normally when you think of a client
-server architecture you'd say,

0:01:34.620000 --> 0:01:40.140000
 oh, okay well if the NMS is running on
 an actual server like a Linux server

0:01:40.140000 --> 0:01:45.960000
 that must be the server component
 of the client-server architecture.

0:01:45.960000 --> 0:01:48.380000
 But in this case it's actually
 just reversed.

0:01:48.380000 --> 0:01:53.760000
 You see in a client-server architecture
 normally we think of the client

0:01:53.760000 --> 0:01:57.720000
 as sending a request
 to the server.

0:01:57.720000 --> 0:02:01.680000
 The server is holding on to some sort
 of data, the client wants it and

0:02:01.680000 --> 0:02:03.580000
 then the server responds back.

0:02:03.580000 --> 0:02:10.460000
 Well in SNMP the SNMP manager is actually
 the one who's sending the request.

0:02:10.460000 --> 0:02:14.680000
 So he is actually the client
 as far as SNMP is concerned.

0:02:14.680000 --> 0:02:19.440000
 So the SNMP manager on which you're running
 SolarWinds or Cacti or whatever

0:02:19.440000 --> 0:02:23.440000
 you're running as a software, that's
 where the request will be formatted

0:02:23.440000 --> 0:02:26.440000
 and sent out. So he's the client.

0:02:26.440000 --> 0:02:28.380000
 So who is actually the server?

0:02:28.380000 --> 0:02:30.860000
 Well the server is the SNMP agent.


0:02:30.860000 --> 0:02:34.140000
 He's the one who's holding on to
 the data that you want to get.

0:02:34.140000 --> 0:02:37.820000
 Now the agent, that's most likely going
 to be your router or your switch.

0:02:37.820000 --> 0:02:44.160000
 So an SNMP agent is actually a software
 sub-component or a piece of code

0:02:44.160000 --> 0:02:46.740000
 that understands SNMP.

0:02:46.740000 --> 0:02:51.540000
 That can take in an SNMP request, get
 the data that request is looking

0:02:51.540000 --> 0:02:55.560000
 for and then send an SNMP
 response back out.

0:02:55.560000 --> 0:03:01.080000
 So the SNMP agent can be almost anything
 that's capable of running SNMP

0:03:01.080000 --> 0:03:04.920000
 software. Now in the context of this
 video we're thinking about networking

0:03:04.920000 --> 0:03:09.340000
 devices. So for example your router
 or your switch, your firewall, they

0:03:09.340000 --> 0:03:11.520000
 would all have SNMP
 agents in there.

0:03:11.520000 --> 0:03:14.360000
 It's not like it's a piece of code
 that you have to specially load in

0:03:14.360000 --> 0:03:20.080000
 there. For example Cisco iOS devices,
 Cisco iOS has SNMP agent functionality

0:03:20.080000 --> 0:03:24.200000
 built into it. You just have to turn
 it on with a couple of commands.

0:03:24.200000 --> 0:03:26.480000
 But other things could
 be SNMP agents as well.

0:03:26.480000 --> 0:03:30.840000
 For example IP phones, even fax machines
 and printers could be queried

0:03:30.840000 --> 0:03:36.640000
 via SNMP if they're running agent
 software within that device.

0:03:36.640000 --> 0:03:42.060000
 Now when the SNMP agent gets a query
 for some information, it has to look

0:03:42.060000 --> 0:03:45.160000
 somewhere to find the information
 and send it back.

0:03:45.160000 --> 0:03:50.600000
 Where is the information
 stored within that agent?

0:03:50.600000 --> 0:03:55.580000
 It's stored in another component called
 the management information base,

0:03:55.580000 --> 0:03:59.200000
 the MIB. So just think of this as being
 a database or as it says here

0:03:59.200000 --> 0:04:03.620000
 a data structure where variables
 are stored on the agent.

0:04:03.620000 --> 0:04:09.040000
 So when the SNMP manager sends a request
 saying hey agent, give me the

0:04:09.040000 --> 0:04:12.980000
 quantity of bytes that have passed through
 your interface fast ethernet

0:04:12.980000 --> 0:04:17.680000
 00, that agent will look in the MIB and
 it will look for something called

0:04:17.680000 --> 0:04:22.420000
 a MIB object that corresponds to that
 data that's being requested and

0:04:22.420000 --> 0:04:23.980000
 then it'll send the data out.

0:04:23.980000 --> 0:04:27.700000
 So here's sort of a graphical representation
 of what that looks like.

0:04:27.700000 --> 0:04:32.300000
 So on the right is where
 we have our SNMP manager.

0:04:32.300000 --> 0:04:34.520000
 So this is actually
 our physical server.

0:04:34.520000 --> 0:04:38.760000
 For example our Linux server or our
 Ubuntu server or whatever happens

0:04:38.760000 --> 0:04:42.640000
 to be. And that's running network
 management station.

0:04:42.640000 --> 0:04:45.040000
 So he's an, he is an NMS.

0:04:45.040000 --> 0:04:47.360000
 So that's the name of his role.

0:04:47.360000 --> 0:04:51.780000
 He is a network management station
 and he's running some sort of SNMP

0:04:51.780000 --> 0:04:57.000000
 software which gives him that
 ability to be an NMS.

0:04:57.000000 --> 0:05:01.300000
 So it's his role to send out SNMP commands
 of which we'll look at some

0:05:01.300000 --> 0:05:03.080000
 of those commands
 in just a moment.

0:05:03.080000 --> 0:05:07.540000
 Now those commands are carried in
 IP and they're carried in UDP.

0:05:07.540000 --> 0:05:09.460000
 So UDP and IP is what
 carries those.

0:05:09.460000 --> 0:05:11.040000
 So it can go across the internet.

0:05:11.040000 --> 0:05:12.980000
 It can go across an intranet.

0:05:12.980000 --> 0:05:15.280000
 Any network that's capable
 of carrying them.

0:05:15.280000 --> 0:05:17.740000
 And those commands are
 going to your agent.

0:05:17.740000 --> 0:05:20.000000
 So your agent could be
 a router, a switch.

0:05:20.000000 --> 0:05:25.200000
 It could be a firewall,
 a printer, an IP phone.

0:05:25.200000 --> 0:05:28.820000
 So that agent has to be running some
 agent software which in the case

0:05:28.820000 --> 0:05:31.560000
 of routers and switches
 is already preloaded.

0:05:31.560000 --> 0:05:34.820000
 And it also has to have a special database
 called a MIB database, Management

0:05:34.820000 --> 0:05:37.100000
 Information Base.

0:05:37.100000 --> 0:05:46.480000
 So just once again an SNMP agent is
 a device running SNMP software that

0:05:46.480000 --> 0:05:50.500000
 contains a MIB. For example here we
 see some routers and switches that

0:05:50.500000 --> 0:05:52.320000
 could be that functionality.

0:05:52.320000 --> 0:05:57.340000
 Now let's just talk about the
 MIB a little bit more here.

0:05:57.340000 --> 0:06:02.100000
 So as I mentioned, MIB stands for
 Management Information Base.

0:06:02.100000 --> 0:06:03.520000
 It's a database.

0:06:03.520000 --> 0:06:08.600000
 It's a database of structured data called
 variables or objects which are

0:06:08.600000 --> 0:06:10.860000
 stored in a hierarchical fashion.

0:06:10.860000 --> 0:06:12.460000
 What the heck does that stand for?


0:06:12.460000 --> 0:06:15.200000
 All right, let's keep
 going on here.

0:06:15.200000 --> 0:06:19.740000
 So if you recall from a previous video
 I gave or videos from other instructors

0:06:19.740000 --> 0:06:24.980000
 or just something you've read, you remember
 that way back in 1988, there

0:06:24.980000 --> 0:06:28.900000
 was an RFC that came out, I think it
 was RFC 1065 or somewhere around

0:06:28.900000 --> 0:06:30.820000
 there. They said, you know what?

0:06:30.820000 --> 0:06:36.700000
 We should sort of conceptualize and
 standardize what devices have inside

0:06:36.700000 --> 0:06:39.620000
 them. For example, devices
 have interfaces.

0:06:39.620000 --> 0:06:42.480000
 Okay, we should have a
 number for interface.

0:06:42.480000 --> 0:06:44.840000
 In what do interfaces store?

0:06:44.840000 --> 0:06:47.760000
 Well, regardless of what device you
 are, an interface typically stores

0:06:47.760000 --> 0:06:52.320000
 counters. All right, let's put a
 number for interface encounters.

0:06:52.320000 --> 0:06:54.140000
 Okay, what do counters
 typically store?

0:06:54.140000 --> 0:06:56.520000
 Well, they typically store
 like bit counts.

0:06:56.520000 --> 0:06:59.740000
 Okay, let's do a number for interface
 counter bit counts.

0:06:59.740000 --> 0:07:01.260000
 They store byte counts.

0:07:01.260000 --> 0:07:03.280000
 They store counters of errors.

0:07:03.280000 --> 0:07:06.440000
 So they said, let's standardize
 all this stuff.

0:07:06.440000 --> 0:07:12.240000
 And then each one of these things is
 accessible in a MIB via this tree.

0:07:12.240000 --> 0:07:16.460000
 Now, as you go along the tree, every
 element in the tree is referenced

0:07:16.460000 --> 0:07:20.960000
 by a number. This is what we call
 the MIB hierarchy or the MIB tree.

0:07:20.960000 --> 0:07:23.600000
 Let me show you an
 example of this.

0:07:23.600000 --> 0:07:26.040000
 So here's like the MIB structure.

0:07:26.040000 --> 0:07:32.180000
 So when an SNMP query goes out, when
 the network management system, the

0:07:32.180000 --> 0:07:37.600000
 NMS send out an SM inquiry, it does
 not say in plain text, if you caught

0:07:37.600000 --> 0:07:40.840000
 it in a stiffer trace or something,
 you wouldn't see in plain text that

0:07:40.840000 --> 0:07:45.680000
 query saying, hey, router five,
 I want to talk to you.

0:07:45.680000 --> 0:07:49.500000
 I want information about all of your
 interfaces, specifically fast, even

0:07:49.500000 --> 0:07:54.680000
 at zero, zero, and the quantity of bytes
 have gone in the outbound direction.

0:07:54.680000 --> 0:07:59.340000
 Now that would be nice if it did it
 in clean clear text or plain text.

0:07:59.340000 --> 0:08:02.660000
 So we could see that in the sniffer trace,
 but that's not the way it works.

0:08:02.660000 --> 0:08:07.760000
 Every one of those things as asking
 for is identified as a MIB object,

0:08:07.760000 --> 0:08:09.780000
 which gives a number to it.

0:08:09.780000 --> 0:08:12.740000
 So for example, you
 can see right here.

0:08:12.740000 --> 0:08:21.120000
 If my NMS was sending out a message,
 an SNMP message, and ultimately that

0:08:21.120000 --> 0:08:25.480000
 message had something to do with a
 Cisco device, and it was trying to

0:08:25.480000 --> 0:08:29.200000
 get some information about interfaces,
 it would have to go through this

0:08:29.200000 --> 0:08:34.260000
 tree. So if we actually caught that message
 in a wire shark sniffer capture,

0:08:34.260000 --> 0:08:37.040000
 we'd see something like this, he'd be
 saying, okay, here's what I'm asking

0:08:37.040000 --> 0:08:45.600000
 for. Dot one, dot three, dot six, dot
 one, dot four, dot one, dot nine,

0:08:45.600000 --> 0:08:47.880000
 dot two, dot two.

0:08:47.880000 --> 0:08:52.580000
 I think that covers all the objects in
 this tree if I didn't miss anything.

0:08:52.580000 --> 0:08:56.580000
 So if it sent out a request, and by
 the way, request is not technically

0:08:56.580000 --> 0:09:01.040000
 the name of an SNMP message, but I'm
 just using it as an example here,

0:09:01.040000 --> 0:09:08.220000
 if you sent out a request for this,
 that is called an object identifier.

0:09:08.220000 --> 0:09:11.040000
 That is stored in our MIB.

0:09:11.040000 --> 0:09:16.740000
 So the SNMP agent, the router, the
 switch, the firewall, this running

0:09:16.740000 --> 0:09:21.920000
 agent software, when he gets a request
 for this object identifier, he

0:09:21.920000 --> 0:09:29.260000
 will go into his MIB, which is just a
 database storing tons of information.

0:09:29.260000 --> 0:09:34.700000
 And he might say, okay, well, in my MIB,
 right here is the location where

0:09:34.700000 --> 0:09:37.400000
 I'm storing this information.

0:09:37.400000 --> 0:09:40.580000
 I'll extract whatever information
 I have from right there.

0:09:40.580000 --> 0:09:45.120000
 If it says interface group, well, we
 haven't gone very deep at this point.

0:09:45.120000 --> 0:09:47.840000
 So it might be, okay, let me tell
 you what interfaces I have.

0:09:47.840000 --> 0:09:51.160000
 Fast Ethan at zero, zero, zero, one,
 one, zero, one, two, and then it

0:09:51.160000 --> 0:09:53.900000
 probably keeps going
 deeper than that.

0:09:53.900000 --> 0:09:57.840000
 So each one of those things, each one
 of those things you could get via

0:09:57.840000 --> 0:10:02.440000
 SNMP is stored in this database called
 a MIB and it's accessible via an

0:10:02.440000 --> 0:10:05.840000
 object identifier, an OID.

0:10:05.840000 --> 0:10:10.320000
 Now, here's the good news.

0:10:10.320000 --> 0:10:16.020000
 99% of the time when you are working with
 SNMP, the NMS that you're working

0:10:16.020000 --> 0:10:22.740000
 on is a GUI based NMS with pull down
 windows and radio buttons and things

0:10:22.740000 --> 0:10:23.980000
 that you can select.

0:10:23.980000 --> 0:10:28.080000
 So you won't even have to
 know what the OIDs are.

0:10:28.080000 --> 0:10:32.060000
 You can just select from a pull down menu
 or something, interface statistics

0:10:32.060000 --> 0:10:36.960000
 or system uptime or routing table.


0:10:36.960000 --> 0:10:41.080000
 And the NMS will know in the background
 what that big long string of numbers

0:10:41.080000 --> 0:10:45.120000
 is, what the correct OIDs for that
 thing that you just selected.

0:10:45.120000 --> 0:10:50.400000
 Now, if you're using a text based or
 a CLI based NMS where you have to

0:10:50.400000 --> 0:10:56.420000
 actually type in an S&P command, then
 yes, that gets very, very painful

0:10:56.420000 --> 0:10:59.580000
 because as part of your command, you
 have to type in this big long string

0:10:59.580000 --> 0:11:03.940000
 of numbers corresponding to the OID
 of what it is that you're trying to

0:11:03.940000 --> 0:11:08.400000
 get. That's why most people, 99% of the
 people would use a graphical based

0:11:08.400000 --> 0:11:16.680000
 S&P system so they don't have to be
 coded it with the MIBS and it has

0:11:16.680000 --> 0:11:19.980000
 in the background of the software what
 all the various numbers are and

0:11:19.980000 --> 0:11:21.440000
 what they correspond to.

0:11:21.440000 --> 0:11:25.560000
 So this is just for your reference so
 you know how in the background how

0:11:25.560000 --> 0:11:30.540000
 SNMP is working between the NMS and the
 agent is querying things for these

0:11:30.540000 --> 0:11:34.420000
 object identifiers referenced
 by these strings of numbers.

0:11:34.420000 --> 0:11:38.780000
 And all these OIDs, which are thousands
 of them, are stored in this big

0:11:38.780000 --> 0:11:43.060000
 database called the Management
 Information Base.

0:11:43.060000 --> 0:11:48.600000
 All right. So talking a little bit more
 here about the network management

0:11:48.600000 --> 0:11:54.960000
 station, we already know this is a
 server running NMS software and it

0:11:54.960000 --> 0:11:56.620000
 can be done in one of two ways.

0:11:56.620000 --> 0:12:01.280000
 Usually what happens is that in the NMS,
 which is a graphical system like

0:12:01.280000 --> 0:12:04.900000
 we see right here, you will set it
 up in advance and you'll say, okay,

0:12:04.900000 --> 0:12:10.020000
 here's the device I
 want you to talk to.

0:12:10.020000 --> 0:12:12.600000
 Of those devices you
 want to reach out to.

0:12:12.600000 --> 0:12:19.420000
 Now your typical routers and switches
 and firewalls by default are not

0:12:19.420000 --> 0:12:21.880000
 enabled for SNMP.

0:12:21.880000 --> 0:12:25.060000
 So if you go to the NMS and you give
 it the IP addresses of your routers

0:12:25.060000 --> 0:12:28.160000
 and switches and firewalls and it starts
 querying them, it's not going

0:12:28.160000 --> 0:12:32.260000
 to get any response back because those
 devices are not listening on UDP

0:12:32.260000 --> 0:12:37.220000
 port 161. So at a very minimum, we have
 to go to our routers and switches

0:12:37.220000 --> 0:12:41.480000
 and firewalls and configure at least
 a couple of commands minimum to get

0:12:41.480000 --> 0:12:44.100000
 them working so they're
 in a listening state.

0:12:44.100000 --> 0:12:49.280000
 So they actually recognize that they
 are supposed to be SNMP agents.

0:12:49.280000 --> 0:12:51.140000
 Okay. So now we've done that.

0:12:51.140000 --> 0:12:54.380000
 We have configured the routers switch
 firewall in advance so it knows

0:12:54.380000 --> 0:12:56.420000
 it's supposed to be an SNMP agent.


0:12:56.420000 --> 0:13:00.360000
 We've configured the NMS and we've
 told them, here's the IP address of

0:13:00.360000 --> 0:13:03.300000
 router one. Here's the IP
 address of switch two.

0:13:03.300000 --> 0:13:04.480000
 So he knows that.

0:13:04.480000 --> 0:13:07.920000
 Now the only thing left to do is you
 have to decide, do you want this

0:13:07.920000 --> 0:13:12.580000
 thing set up to where it pulls
 it on a fairly frequent basis?

0:13:12.580000 --> 0:13:13.940000
 And that's usually what people do.


0:13:13.940000 --> 0:13:17.620000
 They'll have the NMS set up so that
 every five minutes or so, or every

0:13:17.620000 --> 0:13:21.940000
 one minute, whatever you want, it will
 pull the various devices and collect

0:13:21.940000 --> 0:13:23.340000
 these statistics.

0:13:23.340000 --> 0:13:27.020000
 And that way it can start creating
 graphs and bot pie charts.

0:13:27.020000 --> 0:13:30.720000
 And you can see, okay, here's how
 the CPU has looked over time.

0:13:30.720000 --> 0:13:32.480000
 You can click on another
 pie chart.

0:13:32.480000 --> 0:13:36.560000
 All right. Here's how my interfaces have
 looked over time as far as utilization.

0:13:36.560000 --> 0:13:40.880000
 So polling is the most preferential
 and most common way that people set

0:13:40.880000 --> 0:13:45.380000
 up their NMS. But you can also have
 the NMS send out messages manually.

0:13:45.380000 --> 0:13:48.540000
 You could click a button and say, hey,
 send out a request right now to

0:13:48.540000 --> 0:13:51.980000
 router two. I want to get some on
 the spot information right here.

0:13:51.980000 --> 0:13:57.200000
 So it can be done in an unsolicited
 form or it can be triggered.

0:13:57.200000 --> 0:14:01.760000
 The last thing I want to mention here
 on this slide also is that most

0:14:01.760000 --> 0:14:05.960000
 of the time, the way SNMP works is
 that the network management station

0:14:05.960000 --> 0:14:10.000000
 is polling the agent on a
 regular frequent interval.

0:14:10.000000 --> 0:14:14.520000
 But the agent itself, the router, the
 switch could also be configured

0:14:14.520000 --> 0:14:17.560000
 to send what's called
 an SNMP trap.

0:14:17.560000 --> 0:14:21.560000
 Now, what happens with that is if the
 router or switch, if you've configured

0:14:21.560000 --> 0:14:27.180000
 to do this, if a certain trigger happens
 like an interface goes down or

0:14:27.180000 --> 0:14:31.700000
 a routing neighbor comes up or disappears,
 you might say to yourself,

0:14:31.700000 --> 0:14:35.720000
 you know what, if something like that
 happens, that's pretty critical.

0:14:35.720000 --> 0:14:39.640000
 And if my polling is happening every
 five or 10 minutes, I don't want

0:14:39.640000 --> 0:14:43.540000
 to wait for the next polling interval
 to find out about that.

0:14:43.540000 --> 0:14:47.460000
 So you can configure your router or switch
 your SNMP agent to send what's

0:14:47.460000 --> 0:14:52.760000
 called a trap, which means if this event
 happens, if this trigger happens,

0:14:52.760000 --> 0:14:57.520000
 I want you to unsolicited send
 out a trap message to the NMS.

0:14:57.520000 --> 0:15:01.100000
 I want you to let the NMS know,
 hey, this just happened.

0:15:01.100000 --> 0:15:04.220000
 So it's probably going to happen
 in between polling intervals.

0:15:04.220000 --> 0:15:07.560000
 But that's good, because we want the
 NMS to know about that as soon as

0:15:07.560000 --> 0:15:11.460000
 it happens. All right, so let's just
 finish up here really quickly with

0:15:11.460000 --> 0:15:15.240000
 some of the more common
 SNMP message types.

0:15:15.240000 --> 0:15:20.120000
 So I've been using the term an SNMP
 request, but that's not technically

0:15:20.120000 --> 0:15:21.180000
 what they're called.

0:15:21.180000 --> 0:15:24.840000
 So if you're trying to get or request
 some information from a device from

0:15:24.840000 --> 0:15:31.880000
 an agent, technically what you're sending
 is what's called an SNMP polling

0:15:31.880000 --> 0:15:34.360000
 mechanism from the NMS.

0:15:34.360000 --> 0:15:38.200000
 And you're polling a particular
 object ID from the MIB.

0:15:38.200000 --> 0:15:43.560000
 Remember an object ID is that 1.3.6
.4.7, which corresponds to some piece

0:15:43.560000 --> 0:15:48.300000
 of information you're trying
 to get from the agent.

0:15:48.300000 --> 0:15:53.440000
 So this is typically automated to
 occur at predefined intervals.

0:15:53.440000 --> 0:15:59.540000
 Now of less common usage is something
 called an SNMP set message.

0:15:59.540000 --> 0:16:05.780000
 An SNMP set message can actually be used
 to program or configure the device.

0:16:05.780000 --> 0:16:10.580000
 So believe it or not, I could go into
 a router switch, give it basic IP

0:16:10.580000 --> 0:16:15.940000
 connectivity, give it basic routing
 functionality, turn on a couple of

0:16:15.940000 --> 0:16:17.720000
 SNMP commands and walk away.

0:16:17.720000 --> 0:16:18.920000
 Now I haven't done anything else.

0:16:18.920000 --> 0:16:20.440000
 I haven't done any security.

0:16:20.440000 --> 0:16:21.700000
 I haven't done any NAT.

0:16:21.700000 --> 0:16:25.560000
 I haven't done anything but
 basic IP, basic SNMP.

0:16:25.560000 --> 0:16:33.240000
 And now from my server,
 that device remotely.

0:16:33.240000 --> 0:16:39.940000
 Now like I said, most people don't use SNMP
 to modify or change the configuration

0:16:39.940000 --> 0:16:42.880000
 of a device. But you can.

0:16:42.880000 --> 0:16:46.160000
 That's what an SNMP
 set message is for.

0:16:46.160000 --> 0:16:49.260000
 And there's an SNMP response.

0:16:49.260000 --> 0:16:53.660000
 So when you send, when the NMS sends
 out an SNMP get message, hey, I'd

0:16:53.660000 --> 0:16:55.320000
 like to get this OID.

0:16:55.320000 --> 0:16:56.900000
 Please give me some information.

0:16:56.900000 --> 0:17:00.760000
 It will be an SNMP response that comes
 back, contain the information that

0:17:00.760000 --> 0:17:03.540000
 you're looking for.

0:17:03.540000 --> 0:17:07.920000
 I also mentioned something about SNMP
 traps, which is an unsolicited message

0:17:07.920000 --> 0:17:12.000000
 that's sent out by the agent when a
 certain trigger or event happens.

0:17:12.000000 --> 0:17:13.660000
 Now this doesn't happen
 automatically.

0:17:13.660000 --> 0:17:15.800000
 You have to configure
 it to do this.

0:17:15.800000 --> 0:17:19.760000
 Now with an SNMP trap, there
 is no acknowledgement.

0:17:19.760000 --> 0:17:21.920000
 Remember, this is all
 UDP right here.

0:17:21.920000 --> 0:17:33.160000
 So in SNMP to the NMS or not, it had
 no idea if the NMS actually got it.

0:17:33.160000 --> 0:17:35.160000
 It was just unacknowledged.

0:17:35.160000 --> 0:17:38.960000
 Well, with the invention of SNMP version
 three, they said, you know what,

0:17:38.960000 --> 0:17:44.340000
 we should have something similar to a trap
 that the agent sends when something's

0:17:44.340000 --> 0:17:48.740000
 triggered, but that gets
 acknowledged by the NMS.

0:17:48.740000 --> 0:17:53.200000
 And that's what's called an SNMP inform
 only available in version three.

0:17:53.200000 --> 0:17:57.160000
 But now you actually get an acknowledgement
 back from an NMS that he received

0:17:57.160000 --> 0:18:00.880000
 it. So that concludes this video.

0:18:00.880000 --> 0:18:01.580000
 Thank you for watching.

