Documentation
Rate This Document
Findability
Accuracy
Completeness
Readability
SubmitClose

Typical Configuration

Table 1 lists the configuration of each component in the Kubernetes and Docker container scenario.

Table 1 Typical configuration in the Kubernetes and Docker container scenario

Node Type

Typical Configuration

Quantity

Remarks

Management node

Kunpeng server

  • Huawei Kunpeng 920 5220 processors
  • 256 GB or larger memory
  • 2 x 480 GB SATA SSD
  • Avago 3508 RAID controller cards
  • 1 x 10GE SP680 SmartNIC
  • Independent power supply

Three or more
  • Used to deploy core Kubernetes components, including kube-apiserver and kube-controller-manager.
  • Used to deploy HA and load balancing components such as Keepalived, HAProxy, and Nginx.
  • Supports deployment of etcd and SWR.
  • Supports deployment of Elasticsearch and monitoring, log, and alarm components.

Container node

Kunpeng server

  • Huawei Kunpeng 920 7260 processors
  • 256 GB or larger memory
  • 2 x 480 GB SATA SSD
  • Avago 3508 RAID controller cards
  • 2 x 2 x 10GE NIC
  • Independent power supply

Two or more
  • The cluster scale is calculated based on the container scale.
  • 2 x 10GE: storage
  • 2 x 10GE: management + service

Distributed storage node (Ceph)

Hot data: Kunpeng server

  • Huawei Kunpeng 920 5220 processors
  • 128 GB or larger memory
  • 2 x 480 GB SATA SSD
  • 12 x 3.2 TB NVMe SSD
  • 2 x 25GE NIC (LOM) (Every two network ports are bonded together to form a logical port and two logical ports are obtained. The two logical ports are connected to the public network and cluster network respectively.)
  • Independent power supply

Calculated based on the data volume

The calculation formula is as follows: Number of nodes = Planned data volume x 1.5 (Data expansion rate) x 1 (Data compression rate) x 3 (Three copies)/0.8 (Drive utilization)/0.9 (Drive number system conversion)/(12 (Number of drives) x 3.2 TB (Drive capacity))

Warm data: Kunpeng server

  • Huawei Kunpeng 920 5220 processors
  • 192 GB or larger memory
  • 2 x 480 GB SATA SSD
  • 2 x 3.2 TB NVMe SSD for acceleration
  • 12 x 8 TB HDD
  • 2 x 25GE/10GE NIC (LOM) (Every two network ports are bonded together to form a logical port and two logical ports are obtained. The two logical ports are connected to the public network and cluster network respectively.)
  • Independent power supply

Calculated based on the data volume

The calculation formula is as follows: Number of nodes = Planned data volume x 1.5 (Data expansion rate) x 1 (Data compression rate) x 3 (Three copies)/0.8 (Drive utilization)/0.9 (Drive number system conversion)/(12 (Number of drives) x 8 TB (Drive capacity))

Cold data: Kunpeng server

  • Huawei Kunpeng 920 5220 processors
  • 128 GB or larger memory
  • 2 x 480 GB SATA SSD
  • 36 x 8 TB HDD
  • 2 x 25GE/10GE NIC (LOM) (Every two network ports are bonded together to form a logical port and two logical ports are obtained. The two logical ports are connected to the public network and cluster network respectively.)
  • Independent power supply

Calculated based on the data volume

The calculation formula is as follows: Number of nodes = Planned data volume x 1.5 (Data expansion rate) x 1 (Data compression rate) x 3 (Three copies)/0.8 (Drive utilization)/0.9 (Drive number system conversion)/(36 (Number of drives) x 8 TB (Drive capacity))