No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search


To have a better experience, please upgrade your IE browser.


NE40E V800R010C10SPC500 Feature Description - Virtual-Cluster-Chassis 01

This is NE40E V800R010C10SPC500 Feature Description - Virtual-Cluster-Chassis

Rate and give feedback:
Huawei uses machine translation combined with human proofreading to translate this document to different languages in order to help you better understand the content of this document. Note: Even the most advanced machine translation cannot match the quality of professional translators. Huawei shall not bear any responsibility for translation accuracy and it is recommended that you refer to the English document (a link for which has been provided).
Virtual Cluster Control Plane Fundamentals

Virtual Cluster Control Plane Fundamentals

Basic Principles

The control plane of the virtual cluster inherits the mature mechanisms of the physical cluster.

As shown in the following figure, each chassis has two MPUs, the four MPUs each have a switching chip, and the CPU of the MPU and interface are both dual-homed to two switching chips in the same chassis. Cross-connections on the chassis are interconnected through four cables.

The following describes the principles of a virtual cluster control plane using the example given in the preceding figure. Both the establishment process for the control plane and the reliability mechanism are outlined.

  1. Control session establishment process

    When starting a virtual cluster, you need to first create the control plane of the cluster. This is a two-step process:

    • By using the control plane connection, four CPUs on the MPU exchange information and select the master MPU. The master MPU is selected as follows:
      1. By default, the MPU of the chassis with the lowest chassis ID is selected as the master MPU.
      2. If the MPU sends a competition packet but does not receive the response packet from the master MPU within a set time frame, and also does not receive competition packets from the other MPUs, it assumes the role of master MPU.
    • After the master MPU has been successfully selected, the master MPU selects a slave MPU for the system. The slave MPU is selected as follows:
      1. By default, the master MPU of the backup chassis is selected as the slave MPU (the chassis which contains the master MPU is called the master chassis, while the other is called the backup chassis).
      2. If the master MPU does not select a slave MPU in the selection time frame, the slave MPU in the same chassis as the master MPU is selected as the slave MPU.

      The virtual cluster preferentially selects as the slave MPU from a chassis other than the one that contains the master MPU, ensuring that if faults occur on the master chassis, traffic can be switched to the backup chassis, reducing the impact of device faults. When the selected master and slave MPUs are both in the master chassis, the MPU of the backup chassis meets the selection criteria for being the slave MPU and so sends a request packet to the master MPU. After receiving the request packet, the master MPU determines that the MPU of the backup chassis chassis meets the selection criteria for being the slave MPU and it resets the current slave MPU selection, selecting the MPU of the backup chassis as the new slave MPU.

    • Active loop blocking on the control plane

      The previously described four cable connection of the control plane forms a figure-of-eight between the four switching chips in the two chassis. Therefore, during internal cluster communication, a broadcast packet can cause a broadcast storm.

      During the establishment of the cluster, the topology of the control plane must be identified and a control link within it must be blocked (for example, in the preceding figure, link DD between LSW12/LSW22 must be blocked) to prevent a loop from occurring.

  2. Control plane reliability process

    After loops have been prevented on the control plane, there are typically three inter-chassis control links, protecting one another.

    If a fault occurs on one of these control links, the link that was blocked to prevent loops can be restored to maintain three-link mutual protection.

    If faults occur on more than one control link, as long as there is at least one control link that is normal, the inter-chassis communication on the control plane remains normal. For example, if link AA between LSW11/LSW21 is normal, then communication from CPU11 to CPU21/CPU22 can be implemented through link AA, without the communication on the control plane being interrupted.

Updated: 2019-01-03

Document ID: EDOC1100055054

Views: 2777

Downloads: 33

Average rating:
This Document Applies to these Products

Related Version

Related Documents

Previous Next