Syncing Question

[speedrcj]

I've been trying to get rsync to work in sort of a strange way, and I'm curious if anyone has any suggestions...

I have 7 machines, and they each have a directory that I'm trying to use rsync to keep backed up. The problem is, they were previously slightly unorganized, so they have MOSTLY the same content, but not entirely. I want my backup to be a compilation of everything that's in the folders, with no copies. Some of them have different modified times, and created times. When I run rsync 1 computer at a time, it copies all of the files over from the first computer, and then since the files have slightly different metadata, it recognizes them as a different file and copies them again. Unfortunately some of the files are rather large and it's taking quite a bit of time to run. I'm not confident enough to make one master directory and write over all of the individual nodes, so I'm curious if I can have rsync run once, and then check if a file is present by NAME only, and if it is then skip it. Any suggestions? Sorry if this doesn't make much sense. I attempted the --size-only flag, and for some reason it still recognized the files as different among the independant machines.

[bclark4444]

I think i may understand what you are saying, and here's my suggestion: manually run md5sum on both files to verify they are identical. If they are not, then thats the problem.

As a side note: dont do --size-only. Thats just a bad idea, especially for db files or other files that may change internally but remain the same size.

Use --checksum instead

Let me know if you are still having issues.

Thanks
--Brett