Poor performance, new messages not showing up

[ajkessel]

I recently switched to Proton after running my own mail in a colo facility for many years. Bridge performance has been abysmal and I'm wondering if this is normal.

I've tested the production release (3.15.0) as well as building the latest from git master (commit 61ca604), both Linux and Windows builds, on three separate boxes. I cross-built the Windows builds from Linux (GOOS=windows TARGET_OS=windows CC=x86_64-w64-mingw32-gcc).

The problems I'm seeing:

• On both Windows boxes and the Linux box, resyncing from scratch takes about 12 hours. Every time. According to the web UI, I currently have about 16GB of mail on the Proton server. With my normal network bandwidth on gigabit ethernet and fiber, that would download in about 15 minutes. I can understand there is some additional network overhead and CPU involved on both ends, but is it expected that it would take 50 times as long to download as other content?
• After the initial sync occurs, new messages and new folders don't show up. This is true on all three devices and both OSes (Windows and Linux, as well as WSL). This is not a MUA issue -- when I telnet into the Bridge to simulate an IMAP session, it does not have recent messages or recently created folders even when the Bridge CLI reports the sync is complete. In other words, folders and messages that show up through the web UI do not exist on the Bridge (per IMAP query) even after waiting for several hours. If I reset, after about 12 hours, the newer messages do show up, but then newly-arrived messages and newly-created folders after that do not.

One theory I have is that while easy switch is still going, the Bridge is syncing messages in the order they are added to the mail store by easy switch, rather than in the order they are arriving. So it is syncing 10-15 year old messages before new arrivals. If this is the case, the problem might go away after easy switch import is done, but it's been running for about a week now. In any event, it seems like a bug for the bridge to prioritize old mail over new.

I've tried going through front-line technical support but hit a brick wall intended for less sophisticated users who are doing things wrong.

This is not a network or device issue. I've tried with two different ISPs (Verizon FIOS and SBC Global) in multiple US states (Massachusetts, Georgia) as well as over AT&T 5G (same geographic locations). My devices at home consistently get about 1gbps up/down throughput via Ethernet and fiber (no WiFi involved). The main Windows box is a 2024 P1G7, Ultra 9 185H, 2300 Mhz, 16 Cores, 64GB RAM, Nvidia GPU. Whether I'm on or off VPN I get the same behavior.

I have telemetry enabled and am not seeing any errors or warnings in log files.

Happy to supply logs or try patches. This just doesn't seem like acceptable behavior for a production app.

[ajkessel]

Easy Switch has finished and the Bridge is still not functional for me on Windows or Linux. It takes about 48 hours to sync now but inevitably hits an internal error and has to resync from scratch, so I've never had any sustained period of time of actually being able to use it. I've tried reporting to support channels and the bridge email address and gotten nothing back.