1
00:00:00,990 --> 00:00:06,800
You want to use stacking technology you need to buy the right switches so essentially you need to buy

2
00:00:06,800 --> 00:00:10,490
the right product for the feature that you want.

3
00:00:10,520 --> 00:00:17,390
As I mentioned Cisco have supported stacking for a long time so just such as the 30 750 have supported

4
00:00:17,450 --> 00:00:19,480
tech wise for many years.

5
00:00:19,490 --> 00:00:28,270
Yes some examples flics Steck was introduced in 2010 flex stack plus was introduced in 2013.

6
00:00:28,550 --> 00:00:37,030
The switches that support flex stack all the 29 60 yes and 29 60 x 4 flex stack plus you need a 29 60

7
00:00:37,030 --> 00:00:42,930
x or 29 60 x are the speed of a single stack length in both directions.

8
00:00:42,930 --> 00:00:49,010
We're using full duplex is 10 gigabits per second for flex stack and 20 gigabits per second for flex

9
00:00:49,010 --> 00:00:56,760
stack plus the maximum number of switches supported in one stack is full for flex stack and 8 for flex

10
00:00:56,780 --> 00:00:59,600
stack plus in the real world.

11
00:00:59,600 --> 00:01:06,020
Have a look on the Cisco documentation and the data sheets for any switch that you want to buy to ensure

12
00:01:06,020 --> 00:01:10,650
that it supports the speeds and capabilities that you need.

13
00:01:10,670 --> 00:01:17,920
Now Chessie aggregation is another Cisco technology which allows you to make multiple switches operate

14
00:01:17,960 --> 00:01:22,020
as a single switch from a big picture perspective.

15
00:01:22,140 --> 00:01:30,020
In a lot of cases such stacking is used after the Access Layer whereas Shecky aggregation is used for

16
00:01:30,170 --> 00:01:34,930
more powerful switches used in the distribution and core layers.

17
00:01:34,940 --> 00:01:41,480
So in summary Shashi aggregation is used for high end switches as an example Chessie based switches

18
00:01:41,540 --> 00:01:45,400
used in the distribution and code layers of Kempis networks.

19
00:01:45,440 --> 00:01:51,340
It does not require special hardware adapters but rather uses Ethan interfaces on switches.

20
00:01:51,440 --> 00:01:54,520
It typically aggregates only two switches.

21
00:01:54,560 --> 00:01:59,180
It's more complex to configure but provides more options.

22
00:01:59,630 --> 00:02:06,620
Now from a big picture point of view Chessie aggregation is the same as switch stacking multiple switches

23
00:02:06,680 --> 00:02:12,710
act as one switch which gives you both availability and design advantages.

24
00:02:12,710 --> 00:02:20,330
However one of the big reasons for Shishi based aggregation is high availability design's technology

25
00:02:20,330 --> 00:02:27,590
such as Cisco virtual switching system or VSS is supported on the Cisco Sixty five hundred and sixty

26
00:02:27,590 --> 00:02:30,120
800 series switches.

27
00:02:30,140 --> 00:02:32,940
Have a look and Cisco's Web site for more details.

28
00:02:33,230 --> 00:02:35,090
But just a quick overview.

29
00:02:35,360 --> 00:02:41,570
Now even if you're not using Chessie aggregation you need high availability in the core and distribution

30
00:02:41,570 --> 00:02:43,310
layer of your network.

31
00:02:43,950 --> 00:02:50,190
As discussed one of the reasons for having multiple switches in the distribution or Colyer is to provide

32
00:02:50,220 --> 00:02:53,410
redundancy in case one of those switches goes down.

33
00:02:53,670 --> 00:03:01,800
So we use technology such as HAARP spanning tree and others to provide better redundancy and better

34
00:03:01,800 --> 00:03:03,410
scalability.

35
00:03:03,420 --> 00:03:10,500
However the downside is cost you need additional switches and it's also more complex to configure.

36
00:03:10,500 --> 00:03:16,200
You need to think about where you put your spending tree roots as well as your aitches Oppy active Araud

37
00:03:16,230 --> 00:03:25,050
is now Chessie based switch typically has multiple line conc one or more supervised modules and one

38
00:03:25,050 --> 00:03:30,960
or more power supplies for redundancy you want redundant power supplies you want redundant supervisors

39
00:03:31,530 --> 00:03:34,860
and you want a multiple line Conte's in your Chessie.

40
00:03:34,860 --> 00:03:40,950
The idea with supervisor modules is if one of the supervisors goes down the other one can take over

41
00:03:40,950 --> 00:03:47,690
the management of the switch a supervisor a module essentially the brain for the Chessie based switch.

42
00:03:48,090 --> 00:03:51,030
If you lose your supervisor the switch will have no brain.

43
00:03:51,120 --> 00:03:55,030
Hence you have redundant supervisor modules in your switch.

44
00:03:55,320 --> 00:03:59,870
You have redundant power supplies in case there's a problem with one of the power supplies.

45
00:03:59,970 --> 00:04:05,240
And in addition you'll have multiple connections from your access layer to multiple line cards using

46
00:04:05,240 --> 00:04:11,400
linked aggregation to ensure that if one of the line cards goes down the network can continue functioning

47
00:04:11,400 --> 00:04:14,520
using the redundant line card.

48
00:04:14,520 --> 00:04:21,450
Now with Chessie based aggregation what we're doing is taking multiple chassis based switches and using

49
00:04:21,450 --> 00:04:29,850
either relate to through a channel between multiple Chessie based switches to provide better redundancy

50
00:04:30,330 --> 00:04:35,460
and better throughput to the distribution core of the network.

51
00:04:35,460 --> 00:04:42,780
I've discussed that in a lot of detail in the campus videos that make up of course what we can do is

52
00:04:42,780 --> 00:04:49,230
take that a step further and instead of using linked aggregation between the chassis base switches with

53
00:04:49,230 --> 00:04:56,790
spending tree NHS opii we make the chassis based switches appear to be a single switch in the model

54
00:04:56,790 --> 00:05:00,470
on the left the two switches are independent of one another.

55
00:05:00,690 --> 00:05:03,570
They run their own Mac address tables.

56
00:05:03,570 --> 00:05:07,400
They run the own instance of the spending tree.

57
00:05:07,560 --> 00:05:10,900
They essentially act totally independently of one another.

58
00:05:11,070 --> 00:05:16,770
You configure two separate switches in the example and you configure them independently of each other

59
00:05:17,250 --> 00:05:19,310
with an aggregated Chessie environment.

60
00:05:19,310 --> 00:05:24,880
However the switches appear to be one switch to the rest of the network.

61
00:05:24,960 --> 00:05:31,680
You can have multiple physical ports going to different physical switches but you can aggregate them

62
00:05:31,680 --> 00:05:39,510
together using multi Chessie eith a channel because logically a one has two physical connections to

63
00:05:39,510 --> 00:05:46,830
the same switch even though physically its two physical connections to different switches and the different

64
00:05:46,830 --> 00:05:48,650
ways to implement this.

65
00:05:48,690 --> 00:05:55,740
We can use multi Chessie isa channel but use an active standby control plane where one of the payes

66
00:05:56,220 --> 00:05:59,980
acts as the switch for the control plane protocols.

67
00:06:00,000 --> 00:06:05,300
So one of the switches is in control of spending triva ether channel up and running protocols.

68
00:06:05,550 --> 00:06:12,300
But to take advantage of the fording power of the supervisor modules on both switches we have active

69
00:06:12,300 --> 00:06:13,700
active data planes.

70
00:06:13,860 --> 00:06:18,890
We are lead to forwarding and Leia's free forwarding is done by both switches.

71
00:06:18,890 --> 00:06:25,150
The switches synchronize their mac address tables and rotting tables to support this there's a single

72
00:06:25,150 --> 00:06:26,750
switch management plan however.

73
00:06:26,860 --> 00:06:34,120
In other words you manage both switches on the active switch when you change the config of the active

74
00:06:34,120 --> 00:06:34,650
switch.

75
00:06:34,690 --> 00:06:39,880
That configuration is synchronized automatically with the standby switch.

76
00:06:39,880 --> 00:06:47,120
Now you could take us a step further where you have an aggregated virtual switch and an aggregated access

77
00:06:47,120 --> 00:06:48,090
switch.

78
00:06:48,100 --> 00:06:55,210
Physically We've got two switches in the distribution layer not but by using Chessie aggregation and

79
00:06:55,210 --> 00:07:02,920
they appear to be one switch to the Access Layer we have four physical switches but they appear to be

80
00:07:02,920 --> 00:07:11,710
a single virtual switch and then we can run a physical ports in a single ether channel between the distribution

81
00:07:11,710 --> 00:07:13,370
layer and access layer.

82
00:07:13,690 --> 00:07:20,290
And in that case we don't need spanning tree because even though physically we have six switches virtually

83
00:07:20,290 --> 00:07:25,300
We only have two switches with one virtual cable between them.

84
00:07:25,630 --> 00:07:29,710
As always there are caveats and things to be aware of when doing this.

85
00:07:29,710 --> 00:07:36,540
But that's the ultimate vision of a linked aggregation with shishya aggregation and switch stacking.

86
00:07:36,650 --> 00:07:42,580
You're aggregating your physical distribution switches into one virtual switch and you're stacking your

87
00:07:42,580 --> 00:07:49,750
access switches into one virtual switch the simplifies the network because you don't have to worry about

88
00:07:49,780 --> 00:07:53,720
optimizing spanning tree and optimizing HAARP.

89
00:07:53,860 --> 00:07:59,800
There's no need for US IP because we have one virtual aggregates which we would still run spending tree

90
00:07:59,800 --> 00:08:01,310
in case there are problems.

91
00:08:01,540 --> 00:08:06,640
But from a spending tree point of view both of these ports are forwarding because there's only one logical

92
00:08:06,640 --> 00:08:09,490
cable between the two switches.

93
00:08:09,490 --> 00:08:15,080
It really simplifies the management and configuration of a campus network.

94
00:08:15,090 --> 00:08:22,680
So in summary stacking technologies and chessy aggregation technologies allow you to simplify the management

95
00:08:22,740 --> 00:08:27,830
and configuration as well as the forwarding of traffic in an Ethernet network.