Acceleration domain

An acceleration domain refers to a domain that is accelerated by eSurfing Cloud CDN. Acceleration domains can be used for website access, streaming media access, application downloads, and livestreaming.

Edge node

Edge node is a concept introduced for the complex CDN structure. An edge node is a CDN point of presence (POP) used to deliver content to end users through less intermediate nodes. Edge nodes provide lower-latency responses and can connect to end users at a faster speed.

Origin server

An origin server is a self-managed business server and is the source of the content to be delivered.

DNS

Domain Name System is used for domain resolution. DNS resolves domain names to machine-readable IP addresses. Human beings are used to memorizing domain names. However, machines recognize only IP addresses. The process of translating a domain name to an IP address is known as domain resolution. Domain resolution is automatically completed by DNS servers. For example, when a client accesses www.baidu.com, the DNS server automatically translates the domain name to 220.181.112.143. Commonly used DNS service providers include Alibaba Cloud DNS, HiChina DNS, DNSPod, Xinnet DNS, AWS Route53, Dyn, and Cloudflare.

CNAME domain

After you add a domain name to eSurfing Cloud CDN in the CDN console, eSurfing Cloud CDN automatically assigns a CNAME domain, like *.ctadns.cn, to the domain name. You must add a CNAME record to the domain host provided by your DNS service provider to point the domain name to the *.ctadns.cn CNAME domain. After the CNAME record takes effect, requests for accessing the domain name are routed to eSurfing Cloud CDN POPs, and content delivery is accelerated by eSurfing Cloud CDN.

CNAME record

A CNAME record refers to a canonical name record in DNS. A CNAME record is used to resolve a domain name to another domain name (a CNAME domain). The DNS server then resolves the CNAME domain to the IP address of the corresponding server.

Origin host

An origin host determines a specific site running on an origin server to which a back-to-origin request is sent.

1: The origin server is specified by a domain name. For example, if the domain name of the origin server is www.a.com and the origin host is www.b.com, back-to-origin requests are sent to the www.b.com site that runs on the host of the origin server with the IP address resolved from www.a.com.

2: The origin server is specified by an IP address. For example, if the IP address of the origin server is 1.1.1.1 and the origin host is www.b.com, back-to-origin requests are sent to the www.b.com site that runs on the host of the origin server whose IP address is 1.1.1.1.

For more information, see Origin Host.

Follow client request protocol

A CDN POP can follow the protocol of a client request when requesting to the origin server. If the client sends a content access request over HTTPS and when there is a cache miss on CDN, the CDN POP sends a request to origin over the same protocol (HTTPS) to retrieve the content from the origin server. If the client sends a request over HTTP, the CDN POP uses HTTP when retrieving the requested content from the origin server. For more information, see Origin Request Protocol.

National commercial cryptographic algorithms of China

National commercial cryptographic algorithms of China are SSL algorithms recognized by the State Cryptography Administration of China. These algorithms are widely used in HTTPS-based workloads of public service sectors, enterprises, banks, and financial institutions. eSurfing Cloud CDN allows you to install domain name certificates and supports the national commercial cryptographic algorithms of China. This helps meet the requirements of financial institutions, public service sectors, and enterprises for HTTPS-based content delivery acceleration.

For more information, see NCCA-based HTTPS.

Object chunking

After the object chunking feature is enabled, eSurfing Cloud CDN queries the target file from the origin server in chunks of a fixed size regardless of whether the client sends a range request. If the origin server supports range requests, it returns the data within the specified range, and the response does not include the Transfer-Encoding: chunked header. This can improve file delivery efficiency and cache utilization. For more information, see Range Origin Fetching.

Static content

Static content refers to content that remains unchanged regardless of how many times the content is requested by users during a certain period of time. Static content includes HTML files, CSS files, JavaScript files, images, videos, software installation packages, APK files, and compressed files. Static content can be cached on CDN POPs. You can configure the cache TTL for static files in the CDN console.

Dynamic content

Dynamic content refers to content that changes each time it is requested. Dynamic content includes JSP, ASP, PHP, PERL, and CGI files and API operations. Typically, dynamic content appears in database query scenarios in which the data dynamically changes, such as remaining ticket inquiries, payment information management, and dynamic sessions. In general, CDN does not cache dynamic content. You can use Integrated CDN to accelerate dynamic content delivery.

Cache key

A cache key is the unique identifier of a file cached on a CDN POP. Each file cached on a CDN POP has a unique cache key. By default, the cache key of a file is the original URL (including a query string) of the request that is sent from a client to access the file. You can customize the cache key to point different URLs of client requests for the same file to the same cached file. This helps increase the cache hit ratio and reduce the number of back-to-origin requests. For more information, see Customizing Cache Keys.

Ignore query strings when caching

For cacheable files, CDN POPs use the original URLs in user requests as cache keys by default. If the original URLs carry query strings and different parameters in the query strings specify the same file, you can enable the feature for ignoring query strings when caching. After this feature is enabled, CDN POPs use URLs without query strings as cache keys. This helps increase the cache hit ratio and reduce the number of back-to-origin requests. For more information, see Setting the Cache TTL.

Purge

If you want to update the content of a file cached on CDN POPs before the cache TTL expires, submit a purge task in the CDN console. You can specify URLs, directories, or regular expressions for purging. After you submit a purge task, CDN POPs delete the cached contents or mark the cached contents as expired. When an end user requests the content after the content is purged, the corresponding CDN POP retrieves the new content from the origin server, returns the new content to the end user, and then caches the content. The purge feature reduces the cache hit ratio. For more information, see Purge.

Prefetching

After you submit a URL prefetching task, CDN pulls the files from the origin server and caches them on CDN POPs. When end users request the files, CDN POPs directly return the cached files to the users. The prefetching feature helps increase the cache hit ratio and reduce workloads on the origin server. It is suitable for scenarios such as releasing popular files. For more information, see Prefetching.

SSL/TLS

Secure Sockets Layer (SSL) is a secure communication protocol built upon TCP. It enables secure communication and maintains data integrity. Transport Layer Security (TLS) is the standardized version of SSL to secure Internet communication and ensure data integrity. The most typical application of TLS is HTTPS, which is also known as HTTP over TLS. HTTPS is the secure version of HTTP. It runs at the layer between the HTTP layer and the TCP layer to provide data encryption and decryption services for communication at the HTTP layer. For more information, see Configuring HTTPS.

OCSP

Online Certificate Status Protocol (OCSP) is an interface provided by a certificate authority (CA) for querying certificate status. OCSP provides a real-time response mechanism. After the CA server receives a certificate query request from a client, the CA server checks whether the certificate is valid and responds in real time. During the SSL handshake, the client queries OCSP information in real time and blocks the subsequent process before it obtains a query result. This may cause a blank page for a relatively longer period of time in weak network conditions and decrease HTTPS performance, resulting in a poor user experience. The OCSP stapling feature can help mitigate blocking issues during the OCSP query. For more information, see Configuring OCSP Stapling.

QUIC

Quick UDP Internet Connections (QUIC) is an experimental transport layer network protocol. It provides the same security protection level as TLS/SSL and reduces the connection and transmission latencies. QUIC is mainly used for HTTP. HTTP/3 (RFC 9114) built on top of QUIC has all the advantages of HTTP/2 and provides better performance than HTTP/2 in weak network conditions because of the features of QUIC. QUIC was developed by Google. It was released in 2012 and submitted to the Internet Engineering Task Force (IETF) in 2013. In May 2021, IETF published QUIC as RFC 9000. For more information, see HTTP/3 over QUIC.

Back-to-origin rate

The back-to-origin rate is classified into the back-to-origin request ratio and the back-to-origin traffic ratio.

Back-to-origin request ratio: Refers to the percentage of requests for resources that are not cached, have expired, or cannot be cached on CDN POPs to the total requests. Back-to-origin request ratio = Number of back-to-origin requests/Total user requests.

(Note: For domain names for which the object chunking feature is enabled, the number of back-to-origin requests sent by CDN POPs may be greater than the number of CDN access requests sent by users. This is because CDN retrieves data from the origin server in chunks while no chunking is performed when user requests are sent to access CDN. In such a case, the value of the back-to-origin request ratio cannot be used for reference.)

Back-to-origin traffic ratio: Back-to-origin traffic refers to the amount of traffic that is generated when CDN POPs send back-to-origin requests to retrieve content from

origin servers. Back-to-origin traffic ratio = Back-to-origin traffic amount/User request traffic amount. A lower back-to-origin traffic ratio indicates better performance.

Cache hit ratio

The cache hit ratio is classified into the traffic hit ratio and the request hit ratio. A lower CDN cache hit ratio leads to higher workloads on the origin server, thus reducing the static content access efficiency.

Calculation methods: Traffic hit ratio = 1 - Back-to-origin traffic amount/Total traffic amount (data is collected every 5 minutes)

Request hit ratio = 1 - Number of back-to-origin requests/Total number of requests (data is collected every 5 minutes)

You can collect and query statistics about the traffic hit ratio and request hit ratio in the eSurfing Cloud CDN console. For more information, see Querying Usage.

DNS resolution time

The amount of time it takes for the domain name system (DNS) to resolve the specified domain name to the corresponding IP address.

TCP connection time

The amount of time it takes for a browser or client to establish a TCP/IP connection to a web server.

SSL handshake time

The amount of time it takes for a browser or client to establish a Security Sockets Layer (SSL) connection to a web server.

Time to first packet

The amount of time that elapses from the point in time when an HTTP request is sent from a browser or client to the point in time when the browser or client receives the first data packet returned from the web server.

Content download time

The amount of time that elapses from the point in time when the browser or client receives the first data packet returned from the web server to the point in time when all content of the page is received. This metric is used to monitor the content delivery quality on a page.

Content Delivery Network

Terms