ECS Overview

An ECS is a basic computing unit that consists of vCPUs, memory, OS, and EVS disks.

After creating an ECS, you can use it like using your local computer or physical server, ensuring secure, reliable, and efficient computing. ECSs support self-service creation, modification, and operation. You can create an ECS by specifying its vCPUs, memory, OS, and login authentication. After the ECS is created, you can modify its specifications as required. This ensures a reliable, secure, efficient computing environment.

The cloud platform provides multiple ECS types for different computing and storage capabilities. One ECS type provides various flavors with different vCPU and memory configurations for you to select.

For details about ECS types, see ECS Types.
For details about all ECS statuses in a lifecycle, see ECS Lifecycle.
For details about ECS specifications, see A Summary List of x86 ECS Specifications.
For details about how to select ECS specifications, see Selecting ECS Specifications.

Instance Specifications

**Table 1** Instance specifications
Parameter	Description
Processor	The physical CPU model of an ECS. Different processors have different performance.
Basic/Turbo Frequency	Two core performance indicators of a processor. Basic frequency: the frequency at which a processor runs stably and continuously in the default state, which indicates the guaranteed performance. It is also known as the CPU clock speed. Turbo frequency: the peak frequency that a processor can automatically boost its core frequency to under specific conditions, which indicates the burst performance. NOTE: Due to kernel optimization or tool restrictions, the turbo frequency cannot be queried. For details, see the instance specification details.
vCPU	ECS supports hyper-threading, which enables two threads to run concurrently on a single CPU core. Each thread is represented as a virtual CPU (vCPU), and a CPU core contains two vCPUs (logical cores). The number of vCPUs, which is the number of logical cores, is defined by the flavor. If hyper-threading is disabled for the processor of an x86 ECS, the number of vCPUs queried from the ECS is half of the number of vCPUs defined by the flavor. You can query the number of vCPUs of an ECS. For example, in the c7.xlarge.2 flavor, there are 4 vCPUs (logical cores). A 2-core physical CPU contains 4 vCPUs (threads). After hyper-threading is disabled, the number of vCPUs queried from the ECS using the c7.xlarge.2 flavor is 2. For details about hyper-threading, see Enabling or Disabling Hyper-Threading.
Memory	The memory capacity of an ECS.
IPv6	Whether an ECS supports IPv6. NOTE: ECS flavors that support IPv6 vary depending on regions and AZs. Whether a flavor supports IPv6 addresses is subject to the console display after you select a region and an AZ on the Buy ECS page.
CPU Computing Performance	Only credit instances (such as T6) have the following metrics: CPU baseline: When the ECS runs at the CPU baseline, the number of CPU credits that an ECS spends per minute is the same as the number of CPU credits that the ECS earns per minute. Average CPU baseline: When an ECS runs at the CPU baseline, the computing performance of each vCPU is the average CPU baseline, which is calculated using the following formula: Average CPU baseline = CPU baseline/Number of vCPUs
Network Bandwidth	Assured network bandwidth: indicates the initial bandwidth configuration of an ECS. It indicates the maximum transmission bandwidth that can be reached under normal network conditions. Maximum network bandwidth: indicates the maximum transmission bandwidth that can be reached by an ECS when the host network bandwidth is idle. No SLA is committed. NOTE: All the instance specifications are verified in the forwarding-only test environment. In actual service scenarios, the instance performance may vary due to factors such as the instance load type, packet length, persistent and non-persistent connections, image version, and networking model. You are advised to perform a stress test to learn about the instance performance first and select appropriate instance specifications. The maximum bandwidth is the total bandwidth allocated to an ECS. If an ECS has multiple NICs, the sum of the maximum bandwidths allocated to all NICs cannot exceed the maximum bandwidth allocated to the ECS.
Network PPS	Network I/O packets over a private network. The maximum number of packets that the ECS can transmit and receive per second. The unit is packets per second (PPS), which means the number of packets sent and received per second. NOTE: All the instance specifications are verified in the forwarding-only test environment. In actual service scenarios, the instance performance may vary due to factors such as the instance load type, packet length, persistent and non-persistent connections, image version, and networking model. You are advised to perform a stress test to learn about the instance performance first and select appropriate instance specifications. For instructions about how to test PPS, see How Can I Test Network Performance?
Network Connections	The number of valid network connections established between devices (such as servers, clients, and routers) at the same time. The number of network connections of an ECS includes the connections established through TCP and UDP. If your service is sensitive to network concurrency, select instance specifications marked with the number of connections based on service requirements.
Max. NIC Queues	The maximum number of queues supported by an individual NIC of the ECS. More queues usually mean that network data can be distributed and processed more efficiently. This reduces the time for data packets to wait for processing, improves the network PPS and bandwidth performance, and reduces the data packet loss rate and network latency. NOTE: For instructions about how to enable NIC multi-queue, see Enabling NIC Multi-Queue.
Max. NICs	The maximum number of NICs that can be attached to an ECS. NICs include primary and extended NICs. NOTE: A NIC refers to an elastic network interface. You can create and configure network interfaces and attach them to your ECSs for flexible and highly available network configurations. For details, see Elastic Network Interface.
Max. Supplementary NICs	The maximum number of supplementary NICs that can be attached to an ECS. Supplementary network interfaces can be attached to both the primary and extended NICs of an ECS. NOTE: A supplementary NIC is a supplement to NICs. If the number of NICs that can be attached to your ECSs cannot meet your requirements, you can use supplementary NICs. For details, see Supplementary Network Interface. Supplementary NICs are only supported by some new instances. For details, see the specifications details.
EVS Disk Bandwidth	ECSs support instance-level storage I/O isolation. Each ECS enjoys their storage bandwidth and IOPS exclusively. This prevents resource contention during peak hours. If the total performance of EVS disks on an instance exceeds the storage I/O capabilities of the instance flavor, the instance flavor's storage I/O capabilities are used. Assured EVS disk bandwidth: the basic storage bandwidth of the ECS specifications, which is the maximum data transmission bandwidth that can be continuously provided by the ECS under normal network conditions. Maximum EVS disk bandwidth: the maximum storage bandwidth of the ECS specifications, which is the maximum bandwidth that can be reached by the ECS when the network bandwidth of the entire host is idle (other VMs on the host have low requirements on the EVS disk bandwidth). No SLA is committed. NOTE: For details about EVS disk performance, see Disk Types and Performance.
EVS Disk IOPS	ECSs support instance-level storage I/O isolation. Each ECS enjoys their storage bandwidth and IOPS exclusively. This prevents resource contention during peak hours. If the total performance of EVS disks on an instance exceeds the storage I/O capabilities of the instance flavor, the instance flavor's storage I/O capabilities are used. IOPS: the number of read/write operations performed by an EVS disk per second. Assured EVS disk IOPS: the maximum EVS disk IOPS that can be continuously supported by the ECS specifications. Maximum EVS disk IOPS: the maximum EVS disk IOPS supported by the ECS specifications when the host services are idle. No SLA is committed. NOTE: For details about EVS disk performance, see Disk Types and Performance.
Local Disk	Local disks are deployed on the host where the ECS resides. They provide local storage access, high storage IOPS, low latency, high throughput, and high cost-effectiveness. Local disks are suitable for scenarios that require high I/O and fast data exchange and processing. Local disk (GiB) indicates the capacity of a local disk. CAUTION: Local disk data reliability depends on the host reliability, with each disk subject to potential individual faults. Data stored on local disks may be lost. Do not use local disks for long-term storage of service data.
Virtualization	ECS uses virtualization technology to divide a host into multiple VM instances. Each instance has independent CPUs, memory, disk, and network resources. ECS supports the following virtualization technologies: KVM, QingTian, and Xen (discontinued).