London proxy nodes in action: can you beat the evening peak delay?
Last Wednesday night at 8:00 p.m. pinch the clock to measure the ipipgo London residential agent, this time is equivalent to the domestic 3:00 a.m., but the United Kingdom is just after work to brush the peak period of the drama. The average latency of the 5 consecutive measurements with SpeedTest is at187ms-223msIt fluctuates between, the most outrageous being a sudden spike to 300ms once, though it recovered within 10 seconds.
Here's a pitfall to be warned about: when testing, don't pick the"Automatic server selection"! Manually specify the local operator server in order to measure accurately. For example, if you use EE network, choose EE's speed measurement point, same for Vodafone. When I was lazy and used auto-match, I've had cases where the latency data was 40ms falsely lower than the actual value.
test period | minimum delay | Maximum delay | packet loss |
---|---|---|---|
9 a.m. | 162ms | 185ms | 0.3% |
8 p.m. | 187ms | 300ms | 1.8% |
Manchester node metaphysical event: weather affecting delays?
When measuring the nodes in Manchester, I encountered a strange thing: the latency generally increased by about 15% on rainy days. on February 28th, when it rained heavily in Manchester, the latency of ipipgo's residential agent rose from the usual 155ms to 178ms, and the packet loss rate also rose from 0.5% to 1.2%. consulting with the technology they realized that many of the old British houses' broadband lines were on the open wires, and the rainy days might indeed affect the physical connection. Physical connection.
Here's a lesson for you.cold technique: In the ipipgo client, change the "retry interval" from the default 2 seconds to 1.5 seconds, so that when encountering unexpected high latency, the switching line will be faster. The probability of web page loading lag can be reduced by one-third by using this method in Manchester node.
Tale of Two Cities comparison: choose London or Manchester?
From the 20 sets of test data, it appears that the London node is suitable forshort connection service(e.g. API calls), the City node is better suited forlong connectivity scenario(e.g. video streaming). As an example: uploading a 10MB file with a London proxy is 3 seconds faster than City, but City is less likely to buffer when playing a 720p video.
One counter-intuitive finding: although London is physically further away from the country, it's rather faster than Manchester at certain times of the day due to more operator peer points. It's advisable to have both regional agents hooked up in the ipipgo backend, using their homeIntelligent RoutingThe function switches automatically.
Why we recommend ipipgo, these details are not available elsewhere.
1. ExclusiveMunicipal IP library: Their City proxies can be pinpointed down to the borough location, and IPs in boroughs like Trafford and Salford just don't work the same way!
2. Dynamic heartbeat mechanism: Automatically detects carrier signal strength every 5 minutes, much more reliable than the fixed reconnection strategy
3. No play"Fake homes"Set: some agents will use the IP of the server room to disguise the residential, ipipgo each IP can check the home broadband registration records
There was also an egg discovered during last week's testing: in the ipipgo client, enterweatherBeing able to look directly at the real-time weather in the current node city is helpful in predicting network quality.
Frequently Asked Questions QA
Q: What should I do if my UK agent is using the card at night?
A: In the ipipgo account settings, change the "network type" from automatic toLeased LineIt's a little more expensive, but bandwidth is guaranteed for the evening rush hour.
Q: Will using both London and Manchester City IPs conflict?
A: Remember to turn on the router atdual-channel modeThe official website of ipipgo has a detailed tutorial, and the speed can be stacked if you use it well.
Q: Low latency in testing but slow in actual use?
A: 80% is DNS pollution, check it in the advanced settingsForced ECS AnalysisThat's it. It's hidden in the secondary menu of the ipipgo client.
Finally, a big truth: the quality of the UK agent varies, the key has to look at the provider has no local resources. ipipgo has its own base station in Manchester this thing, peers basically can not do. The next test is ready to take you to dig a dig Glasgow node performance, want to see the friends of the comment section knock a 1.