When copying large amounts of data, I feel that rclone sometimes "loops" over the full data set several times, and never ends. I have an example now where I am running a copy command, and in the logfile I can see it assessing whether to copy the same files over again.
What is your rclone version (output from rclone version)
rclone v1.56.0
os/version: ubuntu 20.04 (64 bit)
os/kernel: 5.8.0-1039-azure (x86_64)
os/type: linux
os/arch: amd64
go/version: go1.16.5
go/linking: static
go/tags: none
Which OS you are using and how many bits (eg Windows 7, 64 bit)
Linux 5.8.0-1039-azure #42~20.04.1-Ubuntu SMP
Which cloud storage system are you using? (eg Google Drive)
Dropbox source, Azure Dest
The command you were trying to run (eg rclone copy /tmp remote:tmp)
example of a file being seemingly processed twice extracted from the log below
cat logs/r<redacted>.log |grep "julia/kids/Julia/__Webshare__/08. Archive/2008/J811_dokument/InDesign/J811_54_skonhet/Fonts/Spumoni_screen_fonts"
2021/08/26 00:58:25 DEBUG : julia/kids/Julia/__Webshare__/08. Archive/2008/J811_dokument/InDesign/J811_54_skonhet/Fonts/Spumoni_screen_fonts: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 00:58:25 DEBUG : julia/kids/Julia/__Webshare__/08. Archive/2008/J811_dokument/InDesign/J811_54_skonhet/Fonts/Spumoni_screen_fonts: Unchanged skipping
2021/08/26 13:11:06 DEBUG : julia/kids/Julia/__Webshare__/08. Archive/2008/J811_dokument/InDesign/J811_54_skonhet/Fonts/Spumoni_screen_fonts: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 13:11:06 DEBUG : julia/kids/Julia/__Webshare__/08. Archive/2008/J811_dokument/InDesign/J811_54_skonhet/Fonts/Spumoni_screen_fonts: Unchanged skipping
not really, there are errors, but not on those files.
cat logs/rclone-<redacted>-2021-08-25.log |grep ERROR |grep Attempt
2021/08/26 02:55:50 ERROR : Attempt 1/3 failed with 34 errors and: march failed with 26 error(s): first error: unexpected end of JSON input
2021/08/26 15:08:11 ERROR : Attempt 2/3 failed with 27 errors and: march failed with 16 error(s): first error: invalid character '<' looking for beginning of value
Now this is perhaps interesting, looking at that exact timing of the error, it looks as it starts over the checks..
2021/08/26 15:08:11 DEBUG : julia/MFT/julia/03. Freelance/Jenny Thorell/SNO SMINKSTILEN/Miley Cyrus/J1004_snosminkstilen_mascara_MileyCyrus.jpg.jpg: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 15:08:11 DEBUG : julia/MFT/julia/03. Freelance/Jenny Thorell/SNO SMINKSTILEN/Miley Cyrus/J1004_snosminkstilen_mascara_MileyCyrus.jpg.jpg: Unchanged skipping
2021/08/26 15:08:11 DEBUG : Azure container <redacted>: Waiting for checks to finish
2021/08/26 15:08:11 DEBUG : Azure container <redacted>: Waiting for transfers to finish
2021/08/26 15:08:11 ERROR : Attempt 2/3 failed with 27 errors and: march failed with 16 error(s): first error: invalid character '<' looking for beginning of value
2021/08/26 15:08:12 DEBUG : WmiData.csv: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 15:08:12 DEBUG : WmiData.csv: Unchanged skipping
2021/08/26 15:08:15 DEBUG : ARKIV/Tarzan.tif: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 15:08:15 DEBUG : ARKIV/Tarzan.tif: Unchanged skipping
2021/08/26 15:08:15 DEBUG : ARKIV/LF 1997_03 C1 bw.tif: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 15:08:15 DEBUG : ARKIV/LF 1997_03 C1 bw.tif: Unchanged skipping
2021/08/26 15:08:15 DEBUG : Ateljen_BA6/.apdisk: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 15:08:15 DEBUG : Ateljen_BA6/.apdisk: Unchanged skipping
2021/08/26 15:08:15 DEBUG : Ateljen_BA6/.com.apple.timemachine.supported: Size and modification time the same (differ by 0s, within tolerance 1s)
2021/08/26 15:08:15 DEBUG : Ateljen_BA6/.com.apple.timemachine.supported: Unchanged skipping
If there any errors, then rclone performs a complete retry - it doesn't just try the failing files, that could get quite complex (if not impossible) in some situations.
As a workaround, you could do your large copy with --retries=1, then check and fix the failing files individually, and then verify with a large copy again.
The errors you get look somewhat strange to me.
It would certainly help if you can share the full log - otherwise grep the first 10 lines with errors - they may help us undertand your situation better. It is very difficult (if not impossible) to help you while you are moving our eyes around without letting us move them ourselves.
Yes, I know I am a bit paranoid about redacting stuff and not sharing the full logs (that are also still accumulating), and I realise I may be shooting myself a bit in the foot so to say, so that just makes me appreciate your assistance even more.
ah, so in case I am copying a folder, and there are errors in the copy, it will retry the full copy command 3 times?
That actually explains what I am seeing, so for a job with a large amount of files, I should be setting retries to 1 as you say, and then just retry the other ones with --files-from command.
Correct in simple situations, but it really depends on the errors - suppose it fails while reading a folder with 10 subfolders with 1000 files each... Take a look at the errors and evaluate you situation. Share your errors if in doubt of correct handling.