Downloading folder into empty local location causes numerous false overwrite prompts

[HooHost]

When downloading a remote folder from a Linux server to an empty directory on a local Windows 7 machine, several prompts are triggered to confirm overwriting the existing files.

This happens consistently, and it is NOT a file name case issue. The file exists only once on the remote server. Using the rename option, a bytewise comparison with all files against already downloaded versions shows that they are all identical. There are no symbolic links on the remote server.

Why is this happening? What can be done to stop it? The consequence is that downloads have to be micromanaged and can't be left to transfer in the background. It is a big waste of time and a cause of constant concern regarding consistency (each file MUST be checked manually in case it is actually different).

You can see the prompt in the following screenshots. The file size discrepancy is a result of the file system overhead - the files are identical:

filezilla.suspiciousness.jpg (78.7 KiB) Viewed 1302 times

filezilla.suspiciousness.02.jpg (83.8 KiB) Viewed 1302 times

[boco]

The file size discrepancy is because they are ASCII files, and their line endings differ between Linux and Windows OS.

Please check the log pane at the top while transfers are running. Overwrite prompts for empty target directories are usually triggered by transfer failures and the following retry of the transfer.

[HooHost]

There are like 40,000 files in the transfer - the log window is nowhere near large enough to show them all, and it goes too fast to check while running.

I would agree that it sounds like a failed transfer, except the file that is to be overwritten is actually already completely downloaded and is identical to the one that is prompting to overwrite (this suggests that there was no error in the initial download). Is there some end-of-file-transfer bug that is making Filezilla think there was an error? There is nothing in the failed transfers pane either.

If there's a file-based log output or if one can be configured, I can check that.

[botg]

If you limit concurrent transfers to one, a log scrolling too fast is no longer an issue. If the prompt appears, any preceding error should still be shown.

[HooHost]

It still scrolls, but slower. Anyway I was able to pull this from the log. It does look like it is failing, although only after transferring the entire file. What do these commands mean/how can I stop this connection from being aborted? Thanks!!

Status: Starting download of /public_html/wp-admin/includes/screen.php
Command: PASV
Response: 227 Entering Passive Mode (xxx,xxx,xxx,xxx,47,54)
Command: RETR screen.php
Response: 150-Accepted data connection
Response: 150 6.1 kbytes to download
Error: Disconnected from server: ECONNABORTED - Connection aborted
Error: File transfer failed after transferring 6,489 bytes in 1 second
Status: Disconnected from server
Status: Connecting to xxx.xxx.xxx.xxx:21...
Status: Connection established, waiting for welcome message...
Status: Directory listing of "/public_html/wp-content/plugins/wordfence/crypto/vendor/paragonie/sodium_compat/src/Core32/ChaCha20" successful
Status: Retrieving directory listing of "/public_html/wp-content/plugins/wordfence/crypto/vendor/paragonie/sodium_compat/src/Core32/Curve25519"...
Status: Initializing TLS...
Status: Verifying certificate...
Status: TLS connection established.
Status: Logged in
Status: Starting download of /public_html/wp-admin/includes/screen.php

[botg]

This very much looks like a firewall interfering. Are you by chance using Windows firewall? Try disabling it, it is known to be faulty and breaking FTP for no reason other than shoddy engineering.