I’ve been using Duplicati, and just recently came across Duplicacy. It is a really exciting piece of technology. There are some neat things I’m attempting to do, but I thought I’d first check with more experienced users what I should anticipate in each case.
I have folders on multiple machines that are (instantly, always) kept in sync by rsync. However, file permissions, user/group ownership, and possibly other extended meta-data might be different. For the two scenarios:
- If I start duplicacy on
machine1
,
1a) the initial upload is interrupted, can I resume the upload from machine2
? If so, do I gain by rsync
-ing the .duplicacy
folder to machine2
before resuming?
1b) the initial upload completes, can I do future backups from either machine1
or machine2
with indistinguishable results on the remote storage?
- I’m thinking of speeding up the initial upload by doing it on both machines, each responsible for a subset of
SubFolder
1–4 under myRealRepository
. That is, I set upmachine1
's repository to containSubFolder1
andSubFolder2
, andmachine2
's repository to haveSubFolder3
andSubFolder4
. These are symlinks to the corresponding ones underRealRepository
. Once the upload completes, I do a final round of backup by having all subfolders symlinked in. Based on my understanding of how Duplicacy works, the final upload should be very quick and use up little additional storage — is this correct?