Could I use odrive & amazon to replace CrashPlan?

Hi, I currently use CrashPlan to sync/backup my computer. But it struggles with my 4tb of data.

I’d love to just use the unlimited storage at Amazon but it features no sync capabilities.

That’s when I found odrive.

So, is it realistic to backup all my 4tb of data (music, photos, videos and documents) using odrive and Amazon and ditch CrashPlan?

I have lots of hard drive space so I don’t necessarily need the progressive sync capabilities. I want to have local copies of the photos, music and videos to serve up with my Plex Media server.

Thanks for any thoughts/insights you can provide!

1 Like

Sync naturally wants to keep both the local and remote changes in sync. Syncing 4TB will generate a great deal of unnecessary processing since there will never be any changes to sync.

If you were trying to migrate to the cloud and eliminate local storage, you could UNSYNC to solve the problem. UNSYNC would turn the backup folder into placeholders, eliminating the need for continuous sync processing.

Since you want to keep your local storage, UNSYNC is not an option.

However, there is another option. We are creating a “Backup to odrive” feature optimized for backing up data to the cloud. You will be able to backup your local storage to any cloud storage, including Amazon’s fantastic unlimited storage plan. You will be able to use the same cloud storage for backup, sharing, and primary storage.

“Backup to odrive” is almost ready. It should be available by early July. Please stay tuned for the announcement.

3 Likes

Thanks for the reply. You’re right, it sounds like the backup option is what I’m looking for in terms of media and then progressive sync for documents! Can’t wait - I’ll stay tuned!

@Tony: “Backup to odrive” sounds like a terrifically useful option. Am I correct in assuming that this will be a one way sync; i.e. Sync local to cloud but not vice-versa? If so, it would certainly enable nearly any backup solution to effectively be a cloud storage app. To really be useful, Backup to odrive should support both the Pro and Encryption options.

2 Likes

“backup to odrive” will be one-way (not bi-directional sync). Every new file will be backed up to the designated cloud storage.

Yes, you’ll be able to backup any folder on your computer.

And yes it will support encryption soon after the initial release.

We hope to get this feature out very soon.

2 Likes

@trent7590 I’ve been using oDrive to do this exact same scenario (~13TB of video content) to the Amazon cloud for a while now. While it isn’t the fastest way to backup a media library (a limit of internet bandwidth on my part) it has proven itself recently when one of my striped drives failed and a fresh install was imminent. Honestly, though, when this happened I found it easier to use the Amazon Cloud downloader/uploader to procure my lost files from the cloud and place them exactly where I wanted on said server. But the set-it-forget-it of oDrive is fantastic…never forget to upload/backup a ripped movie/tv show, music file, or photo again. And you can’t beat the price of the Amazon Cloud.

1 Like

Will Backup to ODrive allow setting specific folders to backup? I want to backup my media files, but also run weekly backups of the home directories on my linux system. Those backups are encrypted, compressed, and stored in a specific folder. I’d like to have them uploaded my Amazon Drive and then deleted from my local drive.

Backup to odrive? Wow this sounds awesome. This could eliminate the headache I get when I try to backup my Google Chome sync folder. The Chrome folder is always changing so it never stops syncing. This new feature sounds like I can take weekly snapshots without a 24/7 sync relationship.

Yes, you will be able to pick and choose which folders to backup.

1 Like

Will you support Crashplan or Backblaze in the future? I use the unlimited Amazon Cloud Drive, but I would like to have a redundant cloud backup. Using your upcoming one-way backup to Crashplan or Backblaze would be perfect.

We haven’t had a request for Crashplan support yet, but you can suggest it in the feature requests section if you like.

We support Backblaze B2 in read-only mode, and will support write once they implement a change in their API flow.

@1FIST: Has CrashPlan improved their upload speeds? We tried them in 2012 as a possible backup solution. Their feature set looked good and I recall pricing being exceptional.

I don’t recall what our bandwidth was exactly at the time but is was fairly speedy - in the 1.5 - 2 Gbps range. The first hundred or so MB of data transferred at a rate of 10+ Mbps but CrashPlan quickly throttled uploads to 100-200 Kbps - barely faster than dial-up. Downloads were somewhat speedier but had the same throttling behavior after the first chunks transferred.The prospect of a system backup requiring a month or more to upload was not appealing. If a restore was needed, we wanted the data now rather than in weeks.

Amazon Cloud Drive throttles uploads as well but not nearly as much. We see worst-case uploads of ~3 Mbps with ACD. Too slow for system backups and recovery but workable if you want infrequent access to a pile of data.

Perhaps CrashPlan has improved. I’d investigate first with a reasonable (i.e.10’s of GB) data set.

@Tony: Is odrive trying to minimize Class B and C transactions with Backblaze B2? A cloud-based system backup utility we use proved immensly costly paired with B2 due to spawning Class C transactions at a rate far higher than they did with S3.

We aren’t doing anything special with B2 at the moment. Progressive sync naturally helps this, as it reduces the scope of calls to B2. This is assuming there is a virtual folder structure being used where filename prefixes can be used to filter file list calls.

If your B2 account has no virtual directory prefixing, however, I could see the calls adding up as periodic queries to B2 for the content listing would end up listing every single thing in the bucket every time a list of the bucket was asked for. Since the max files returned in a call is 1000, this could add up quick if you have a massive bucket with no virtual (prefixed) structure.