What are the deployment principles of HDFS?

3

NameNodes and ZooKeeper Failover Controllers (ZKFCs) are deployed on two control nodes that work in active/standby mode. JournalNodes are deployed on at least three nodes and each node reserves a copy of backup data. It is recommended that at least three DataNodes be deployed on data nodes.

Other related questions:
Deployment principle of HDFS in a FusionInsight cluster
NameNodes and ZooKeeper Failover Controllers (ZKFCs) are deployed on two control nodes that work in active/standby mode. JournalNodes are deployed on at least three nodes and each node reserves a copy of backup data. It is recommended that at least three DataNodes be deployed on data nodes.

Principles of HSB
The AR supports the HSB function. HSB implementation involves data synchronization and traffic switching. Data synchronization is performed to ensure consistent information on the master and backup devices when the two devices are working normally. Traffic switching is performed to ensure non-stop service forwarding when the master device fails or recovers. The principle for data synchronization is to establish active and standby channels between devices that back up each other. Session entries of the master device can be synchronized to the backup device through the channel at one time, in real time, or periodically. The principle for traffic switching is based on negotiation between the master device and the backup device using VRRP. When the master device fails, a new master device is elected based on VRRP priorities and the traffic is switched to the master device. For details, see “HSB Configuration�?in AR100&AR120&AR150&AR160&AR200&AR1200&AR2200&AR3200&AR3600 V200R008 CLI-based Configuration Guide - Reliability.

What is the working principle of Streaming?
A Streaming topology can be developed and compressed into a JAR file by using Storm APIs or referring to the CQL syntax manual. After the topology is submitted to the Streaming cluster, Nimbus automatically assigns it to the Supervisor node for processing. The topology logs can be viewed on the UI through the port provided by Logviewer.

If you have more questions, you can seek help from following ways:
To iKnow To Live Chat
Scroll to top