1
00:00:00,300 --> 00:00:06,420
For example, if the NMS is pinging a router to see if it's up or down and a response doesn't come

2
00:00:06,420 --> 00:00:11,580
back, the NMS doesn't really know whether that device is down

3
00:00:12,150 --> 00:00:14,710
all it knows is that the response didn't come back.

4
00:00:15,330 --> 00:00:20,880
The device could be down, but there could also be a routing problem or an intermediate device could

5
00:00:20,880 --> 00:00:26,580
be down or a device between the NMS and the router could be down.

6
00:00:27,270 --> 00:00:35,100
So some device along the path could be down or traffic could be blocked by an ACL or a firewall rule.

7
00:00:35,760 --> 00:00:39,900
All the NMS knows is that it can't reach the device that it's pinging.

8
00:00:40,410 --> 00:00:43,920
SNMP on the other hand is different, with SNMP

9
00:00:44,100 --> 00:00:52,660
the NMS is querying a device and the device is telling the NMS via SNMP the state of an interface.

10
00:00:53,220 --> 00:00:58,560
So if a device indicates that interface, FastEthernet 00 as an example is down

11
00:00:59,010 --> 00:01:05,250
the NMS knows 100 percent that that information is accurate because it received that information

12
00:01:05,250 --> 00:01:08,690
via SNMP from the device that it was querying.

13
00:01:09,150 --> 00:01:15,510
Now, in the event that the actual device is down rather than an interface on the device, you'll get

14
00:01:15,510 --> 00:01:16,550
an unknown state.

15
00:01:17,040 --> 00:01:23,640
That's a different type of signifier than if the interface is down on a device that's up and running

16
00:01:24,060 --> 00:01:31,770
in terms of performance monitoring, ICMP or PING is used for availability calculations and latency

17
00:01:31,770 --> 00:01:40,170
response time as SNMP is used for almost all the other statistics, including CPU, memory, buffers,

18
00:01:40,650 --> 00:01:45,690
interface traffic, and errors, and many, many other statistics.

19
00:01:46,380 --> 00:01:55,770
In some cases, for Windows systems, the NMS might leverage WMI or Windows management instrumentation

20
00:01:56,310 --> 00:01:58,860
to check for performance counter type values.