Data center switch core destacking

2. Generally, the stacking function is used in the convergence layer.. The core layer only does high-speed data forwarding and link rendancy
stacking is commonly used in the following situations: first, when the number of campus network information points increases and the network scale expands, in this case, the access layer switch needs higher backplane bandwidth. In the past, several or more gigabytes of stacking bandwidth can not meet the needs of the rapid development of network traffic< Second, the extension of distance. In a switch stack, the distance between several stackable switches is only a few meters. This limits stacking to a small area
Third, function expansion. Traditionally, stacking is a way to connect switches

since the whole stack is made up of multiple member switches stacked by stacking moles and cables, it is more prone to failure than a single device. To ensure high availability, it is necessary to ensure that the whole stacking system can still work normally when a single device or a stacking cable fails. At the same time, it can ensure the seamless upgrade and expansion of the system.

3. blockchain technology is also known as distributed ledger. Generally speaking, if we assume that the database is an ledger, reading and writing the database can be regarded as a kind of bookkeeping behavior. The principle of blockchain technology is to find the fastest and best person for bookkeeping in a period of time, and then send this page of information of the ledger to everyone else in the whole system. This is equivalent to changing all the records in the database and sending them to every other node in the whole network, which can not only ensure the data security, but also ensure the authenticity of the data, so as to enhance the social credit

the trust of blockchain comes from the underlying technology, that is, to exchange historical information for current trust. This is a low-cost credit mechanism. From its birth to its slow growth, that is, blockchain, it is bound to get attention and promotion.

4. There are several in Luzhou trade city and Luzhou high tech Zone. The one in the high tech Zone is called Sichuan wecan Network Technology Co., Ltd. It is doing distributed storage. You can learn about it on the Internet.

5. Bitcoin is a kind of "currency"
"distributed bookkeeping" is just a name compared with "central bookkeeping" of banks
distributed accounting is based on blockchain technology
sometimes the distributed ledger technology refers to the blockchain technology. But the two are not equal
for example, distributed bookkeeping can also be realized through 1.3 billion Chinese people's manual bookkeeping.

6. The function of the uplink port and 1 port are the same! Therefore, it must be noted that in the stage connection, you must not use the uplink and port 1 at the same time. In the stage connection, for example, if you have three switches (the first one is connected to the second one, and the second one is connected to the third one), then in the stage connection, if the first one uses the uplink or port 1, then in the second one, you have to plug the network cable into the port other than the uplink and port 1, and the third one is the same with the second one
in a word, two connected switches can't use uplink and 1 port at the same time!

7. The difference between stack and cascade
1 requires different devices. Cascading can be completed by a twisted pair between switches of any network equipment manufacturer or between switches and hubs. Stacking can only be realized between the devices of our own manufacturer, and the switch must have stacking function. 2. Different requirements for connecting medium. When cascading, only one jumper is needed, while stacking requires special stacking mole and stacking cable. Of course, stacking mole needs to be ordered separately. 3. The maximum number of connections is different. In theory, there is no limit on the number of cascades. However, the number of switches that can be held in the stack will be clearly limited by each manufacturer. 4. Different management methods. The stacked switches are logically a managed device, which can uniformly configure and manage all switches. The cascaded switches are logically independent, so they must be configured and managed in turn. 5. The connection bandwidth between devices is different. When multiple switches are cascaded, there will be cascading bottleneck, which will lead to large forwarding delay. For example, when four 100m switches are cascaded by jumpers, the connection bandwidth between them is also 100Mbps. When computers connected to different switches communicate with each other, they can only communicate through this 100 megabit connection, which becomes the bottleneck of transmission. At the same time, with the increase of forwarding times, the network delay will become very large. When the four switches are stacked together, the stacked cables can provide more than 1Gbps backplane bandwidth, which can realize the high-speed connection between all switches. Although link aggregation technology can be used to increase bandwidth between switches in cascading, it is at the expense of available ports. 6. Different network coverage. Switches can expand network coverage by cascading. For example, take the twisted pair network as an example, the diameter of the network covered by one switch is 100m, the diameter of the network covered by two switches cascaded is 300m, and the diameter of the network covered by three switches cascaded is 400m. The stacking cable is usually only 0.5 ~ 1m, which can only meet the needs of interconnection between switches and will not affect the network coverage.

8. This is a data center switch, which supports virtualization, that is, VSU. Configuration reference
core switch 1:

Switch1? Configuration terminal
enter configuration commands, one per line. End with CNTL/Z.
Switch1(config)# switch virtual domain 1
Switch1(config-vs-domain)# switch 1
Switch1(config-vs-domain)# switch 1 priority 200 ------> The default priority is 100, which is configured as a higher priority. After the VSU is successfully established, it will become the management host< br />Switch1(config-vs-domain)# exit
Switch1(config)# vsl-aggregateport 1 ------> At least two VSL links are needed, and the reliability of one link is low. When link oscillation occurs, VSU will be very unstable< br />Switch1(config-vsl-ap-1)# port-member interface TenGigabitEthernet 1/1 ------> Configure VSL link, heartbeat link and traffic channel between VSU active and standby cores
Switch1 (config-vsl-ap-1) # port member interface tengigabit Ethernet 1 / 2
Switch1 (config-vsl-ap-1) # exit

core switch 2:

switch2 # configuration terminal
enter configuration commands, one per line. End with CNTL/Z.
Switch2(config)# switch virtual domain 1 ------> The domain ID must be the same as the first one
switch2 (config vs domain) # switch 2 ----- & gt; The second device must change its ID to 2
switch2 (config vs domain) # switch 2 priority 150
switch2 (config vs domain) # exit
switch2 (config) # VSL aggregateport 1 --- & gt; At least two VSL links are needed, and the reliability of one link is low. When link oscillation occurs, VSU will be very unstable< br />Switch2(config-vsl-ap-1)# port-member interface TenGigabitEtherne 1/1 ------> Configure VSL link, heartbeat link and traffic channel between VSU active and standby cores
switch2 (config-vsl-ap-1) # port member interface tengigabit Ethernet 1 / 2
switch2 (config-vsl-ap-1) # exit

9. Hua 23 33 53 can be stacked, which has just been confirmed as a design defect

10. If the switch has an uplink port, this port of the next switch is connected to other ports of the previous switch. If the switch does not have a cascade port, it can be connected one by one. Note that the number of cascade layers is limited, and it is not good to have more than four