Linux Google Drive Client

[Chaos6]

With the price drop yesterday, I'm considering switching to Google Drive from my current backup solution (Zoolz). I already pay for a (grandfathered in) $100/yr 400GB plan, so another $20/yr for 1TB seems like a no brainer.

I tried using Google Drive before on a Windows 7 machine to back up my NAS/home server, but the performance was pretty terrible. It would re-scan the entire drive every time the machine came back from sleep. I switched to Zoolz about 6 months ago, and have been generally satisfied with the client (which unfortunately is still Windows based). It actually listens for filesystem changes, so rescans are rare. The downside is that upload speeds are capped at 2Mbps, which is annoying since I have a 35Mbps upload with fios.

I'd love to switch back to Google Drive, but after my previous experience with using the Windows client to sync a network share, I think only a native linux solution is viable. My home server is running Ubuntu Server 12.04 (LTS).

I see there are a few commercial options since I last checked. insync-headless looks promising for a $15 one time fee. Does anyone have experience using Google Drive on a linux machine?

Alternatively, does anyone know of any open source tools that provide an easy implementation of the Google Drive APIs? My backup partition on the server is btrfs, so I probably wouldn't mind writing my own differential sync based on the nightly snapshots as long as the API is pretty easy to deal with.

 

[stevech]

With the price drop yesterday, I'm considering switching to Google Drive from my current backup solution (Zoolz). I already pay for a (grandfathered in) $100/yr 400GB plan, so another $20/yr for 1TB seems like a no brainer.

I tried using Google Drive before on a Windows 7 machine to back up my NAS/home server, but the performance was pretty terrible. It would re-scan the entire drive every time the machine came back from sleep. I switched to Zoolz about 6 months ago, and have been generally satisfied with the client (which unfortunately is still Windows based). It actually listens for filesystem changes, so rescans are rare. The downside is that upload speeds are capped at 2Mbps, which is annoying since I have a 35Mbps upload with fios.

I'd love to switch back to Google Drive, but after my previous experience with using the Windows client to sync a network share, I think only a native linux solution is viable. My home server is running Ubuntu Server 12.04 (LTS).

I see there are a few commercial options since I last checked. insync-headless looks promising for a $15 one time fee. Does anyone have experience using Google Drive on a linux machine?

Alternatively, does anyone know of any open source tools that provide an easy implementation of the Google Drive APIs? My backup partition on the server is btrfs, so I probably wouldn't mind writing my own differential sync based on the nightly snapshots as long as the API is pretty easy to deal with.

For me, Google Drive is not viable - because it's concept is that you put files (folders too?) in a google drive folder on your PC. Their software synchs that one folder to google drive. Your description suggests they have changed?

 

[Chaos6]

Nope, that's still the basic concept for the Google provided clients as far as I know. For me that wasn't really a problem, since I'm actually having it sync from my on-site backup. Getting everything I need into the appropriate "Google Drive" folder is just a matter of mounting the backup partition in the right place, or throwing a symlink in there. There isn't anything I want to put in Google Drive that I don't also want backed up locally.

That said, I believe some of the third party options that use their API have a lot more flexibility than the out of the box client.

 

[stevech]

My ISP uplink speed is only 1Mbps so I don't use any on-line storage for big files.
And I don't upload anything remotely sensitive.
I've been using OpenDrive for a long time. Last 2 years it has finally become quite stable.
My data is just personal stuff .. photos, etc. No financial data goes out.

If I were doing this professionally, I'd probably use Amazon S3 or the Rackspace front end. For reliability and longevity.

 

[Chaos6]

I think I disagree with that philosophy. Just because my data is personal doesn't mean it isn't sensitive. I don't want anyone who I haven't given access to be able to see pictures of my kids any more than I want them to have my tax returns.

That's one nice thing about Zoolz (which is a reseller of Amazon glacier). Nothing can be accessed without my encryption key - even with access to my account.

I remember looking at OpenDrive a while ago, but I was put off after reading their BBB responses (their responses were the disturbing part, not the complaints).

I'd like to try Google Drive out if I can find a decent Linux client - if for no other reason than the fact I already pay for almost as much space as I would need to make it my primary backup solution. While I don't particularly trust Google with unencrypted data - it's a large enough company that I know they have a strong financial incentive to at least secure my data. If I end up moving to it over Zoolz permanently, I'll probably use client side encryption.