Help applying erasure coding to existing storage

Hello,

I’m looking for advice about how to add erasure coding to existing storage. My current setup, using Duplicacy Web:

Storage (not initialized with erasure coding):

  • LOCAL: Storage repository on local HDD
  • B2: Storage repository on BackBlaze B2

Schedules:

  • HOME: Backup contents of home directory on local SSD → LOCAL
  • BULK: Backup large files from local HDD (excluding LOCAL repository) → B2
  • COPY: Copy LOCAL → B2

Assuming Backblaze’s own erasure coding scheme makes Duplicacy’s erasure coding unnecessary on my B2 repository, what’s the best way to replace my LOCAL repo with one that uses erasure coding without disruption to my COPY schedule? I’m hoping that subsequent copies from the new erasure-encoded LOCAL repo to the existing B2 repo won’t require re-uploading to B2 everything from the new LOCAL repo that was already copied there from the existing LOCAL repo.

Thanks!

Luke

You can create a new local storage, let’s call it LOCAL2, that is copy-compatible with LOCAL and with erasure coding enabled. Then run a copy job to copy all backups from LOCAL to LOCAL2. After that, if you don’t want to change your existing schedules, you can just delete LOCAL from the storage page, and then create a new storage named LOCAL that points to the same storage url as LOCAL2. Everything should work the same way as before.

1 Like

I’m also wondering what the easiest way of turning on erasure coding is on my existing storage. But my storage is remote and almost 4 TB so downloading and re-uploading is not an option. I’m therefore trying to understand what exactly the problem is with just starting to back up with erasure coding. @gchen wrote earlier

But what does that mean? What kind of messiness are we talking about?