Features

KMR provides rich management functions and convenient application programing interfaces, so you can perform the data processing and analysis efficiently and automatically, which will save the management cost and use cost:

Elastic expansion

KMR Cluster provides outstanding horizontal expansion. You can increase or decrease nodes flexibly according to business needs, start and stop nodes at any time, and expand the capacity flexibly. It supports daily billing, and supports elastic capacity increase and decrease to adapt to changing business scenarios and save the cluster usage cost.

Standard storage service access

KMR can access to standard storage services (KS3) directly via internal high-speed network. To perform the data processing, you can consolidate original data to KS3 first. Jobs running in KMR Cluster such as MapReduce, Hive, Pig, Spark, etc. can call this data stored in KS3 for the computing, and writes the results back to KS3. KS3 offers lower use cost and high data reliability, and guarantees that original data and computing results can be stored persistently when the cluster is released.

Remote cluster management

Each KMR Cluster can be configured with a public network EIP, which is bound automatically to the cloud host where the master node is located. You can connect to the cluster remotely with SSH tool to customize the cluster configuration or issue computing jobs.

Secure access with SSH key

KMR offers very high network security, and only necessary service ports are opened. In internal network or external network, you can only access to the cluster node through SSH key. Before using SSH to manage the cluster, you need to upload public key of local SSH client to KMR Cluster to generate a key pair, then you can access to the cluster securely. You can manage your own SSH key in the console and add or delete SSH key for different KMR Cluster.

Cluster log consolidation

Usually, KMR Cluster and job logs are distributed among various nodes, and there are different log paths, so it is inconvenient to view them. Moreover, after the cluster is released, you cannot view the log, cannot locate issues and fine-tune the cluster. KMR can store the cluster and job logs in the designated directory on KS3 for easy management and persistent storage.