Staff:
Following up on your explanation comment, I’ve done some speed checking using servers from different locations and my conclusion is that from my location in the SF Bay Area the two new Los Angeles servers are about 50% slower than the remaining Fremont server (Aquila). With Aquila I get about 180 to 190 Mbps down; with Teegarden and Groombridge it’s about 90 to 95 Mbps down. With no VPN I test at 209 Mpbs down and 6 Mbps up. I used Ookla’s speedtest.net for this. If you can suggest a better test site I’ll use it.
My practice has been to use servers from my three closest locations: Fremont, Los Angeles and Phoenix. The Phoenix location, with five servers, has tended to be the slowest; I only use them when Heze is down AND to connect to one of my daily go-to sites--Macintouch.com--which blocks Persei and Aquila. (Oddly, yesterday and today the Phoenix servers have been blazingly fast, giving me test speeds equal to Aquila (around 190 Mbps down. I don’t know what to make of that; I’m assuming it’s an anomaly.)
My internet provider is Comcast, and because I suspect it throttles torrent connections I generally use the TCP 443 protocol. However, for the speed tests I just did I used the faster UDP 443 protocol. (One thing that puzzles me is the in all VPN test cases my upload speed was only about 2 mbps.)
You said the new Los Angeles servers are “quite near in terms of network distance and very near geographically.” I can’t comment on network distance, but I will say while LA and SF/Fremont may APPEAR to be close on a small scale map, in reality they are about 350 miles apart, as the crow flies. That’s a significant distance, even for California (similar to the distance between Rome and Marseilles, France).
I’m not trying to be argumentative but I feel AirVPN has degraded the service of all of your Bay Area subscribers by moving the two Fremont (i.e., Silicon Valley) servers to southern California. I hope you’ll reconsider.