Internet Access Performance Monitoring displays the Internet access event and average latency from each Alibaba Cloud region to various off-cloud regions. Based on this, you can select an appropriate region to deploy your services and improve the access experience for off-cloud users.
Scenarios
A user deploys resources in an Alibaba Cloud region to provide services to Internet end users in neighboring countries. When multiple end users from a country report poor network quality, the user can:
Use View the performance dashboard to obtain information such as the specific network event, occurrence time, and duration.
Use View Internet coverage latency to obtain the latency from Alibaba Cloud resources to each ISP in the target region. Based on the round-trip time (RTT) result range, we recommend the following applications:
RTT<=10ms: Industrial automation, remote driving, ultrasonic testing
RTT<=50ms: Real-time competitive gaming
RTT<=100ms: Real-time interactive gaming, interactive voice
RTT<=150ms: Real-time interactive video
RTT<=300ms: General applications
RTT>300ms: We recommend that you assess whether this latency meets your requirements based on your specific business scenario.
Usage disclaimer
The network latency and other performance data provided by Internet Access Performance Monitoring are for reference only. Alibaba Cloud makes no promises or guarantees regarding the accuracy or real-time nature of this data. Actual network performance depends on real-world conditions.
The Internet access events and Internet access latency displayed in Internet Access Performance come from an ECS probe cluster that Alibaba Cloud builds independently. The probe cluster periodically sends and receives performance probe packets over the network, and then sends the aggregated and summarized data to be displayed on the performance monitoring page. Because the test machines in the probe cluster are completely independent from tenants, they do not affect or consume any of your resources. The performance monitoring data is unrelated to the resources you actually deploy on Alibaba Cloud.
How it works
Geolocation accuracy: For IP location information and ASN attribution information, the IP geolocation data provided by IPIP.NET is used.
IPv4 and IPv6: Currently, network health is monitored only over IPv4.
Data sources: All data is generated based on active probing of the Internet.
Data update frequency: Internet access event data is updated every 15 minutes.
Number of monitored ISP networks: Alibaba Cloud tracks the network quality and traffic volume between each Region and "each country + ISP" worldwide, and continuously tracks target networks with traffic volume to build a quality baseline. Currently, granular data at the province level/state level + ISP is provided for the China/United States regions.
Which network access events are created: Global probe data detects latency or packet loss rate anomalies on the Internet, and aggregates them into Internet events by the dimensions of "country + (province level/state level) + ISP". The map displays Internet events over the past 48 hours that affect the access experience of cloud customers. If the event type is "latency", you can view the start time, duration, event status, and affected location at the event point.
Events are divided into three types by status:
active availability event (packet loss): Within the past 48 hours, compared with the historical quality baseline, the network packet loss event status is still "active" and not recovered.
active performance event (latency): Within the past 48 hours, compared with the historical quality baseline, the network latency event status is still "active" and not recovered.
recovered event: Network events that changed from "active" to recovered status within the past 48 hours.
Internet coverage latency data: Alibaba Cloud tracks the network latency between each Region and "each country + ISP" worldwide at the minute level, and aggregates it into the average network latency over the past 48 hours by the dimensions of "country + (province level/state level) + ISP".
View the performance dashboard
Go to the Internet Access Performance page.
In the Performance Dashboard area, select Source Region from the drop-down list, for example, Singapore.
The map on the left displays the latency from Source Region to each region of Global (country dimension) or China (province level dimension), as well as the locations of network events that occur. When you hover the mouse over any position on the map on the left, you can view the latency data for accessing the region at the mouse position from Source Region.
The right side displays all Internet access events that occurred during probing between Source Region and other regions over the past 48 hours, for example, Latency Degradation or Packet Loss Degradation.
View Internet coverage latency
Go to the Internet Access Performance page.
In the Internet Coverage Latency area, select Source Region from the drop-down list, for example, Singapore.
Filter by Destination Country, Destination Province, and Destination ISP as needed to view the target latency.
More information
Regions that support Internet access performance
Area | Regions supporting Internet Access Performance Monitoring |
China | China (Qingdao), China (Beijing), China (Zhangjiakou), China (Hohhot), China (Ulanqab), China (Hangzhou), China (Shanghai), China (Shenzhen), China (Heyuan), China (Guangzhou), China (Chengdu), China (Hong Kong) |
Asia Pacific Others | Japan (Tokyo), South Korea (Seoul), Singapore, Malaysia (Kuala Lumpur), Indonesia (Jakarta), Philippines (Manila), Thailand (Bangkok) |
Europe and Americas | US (Silicon Valley), US (Virginia), Germany (Frankfurt), UK (London), Mexico |
Middle East | UAE (Dubai), SAU(Riyadh) |
Term descriptions
RTT (Round-Trip Time): The total time it takes for a data packet to travel from the sender to the receiver and back to the sender, that is, the round-trip time for "sending out and receiving back". For example, the duration you get when pinging a website is the two-way (round-trip) latency. Generally, a shorter latency indicates a shorter network path for accessing the service and a better experience.
packet loss rate: The proportion of data packets that fail to reach the destination to the total number of data packets sent during network transmission, usually expressed as a percentage (%). According to historical experience data, the Internet packet loss rate is strongly correlated with user satisfaction: as the packet loss rate rises, the available bandwidth at the application layer drops sharply. Therefore, the lower the packet loss rate the better overall, and we recommend keeping it within 5%.