Google Drive is a cloud based storage facility allowing you to store your documents, binaries, photos: in fact any sort of data.
The recent and long overdue price reductions of their storage plans (1TB is now 10USD per month) has made me review my storage policies
Reliability/ Functional Issues
Aug 2013 Drive Issues
Oct 2013 Insync Saves the Day
Google Drive could be better! For moderate users (and this used to be Marcus also) it seems to work 100%. However things can turn ugly (in my experience) when
a) You store >100,000 files and over 100GB storage
b) You need to move the drive or resync the drive without fully uploading or downloading everything again
Follow the posts for more details.
Let's just restate that for moderate users, you will not have a problem. But since I experienced missing files/ directories on servers with Terabytes of storage and several millions of files, many of which change on a daily basis, I now use the Google Drive as a Cloud Accessible secondary storage device. Let's see how:
Current Marcus Solution
A normal Google Drive installation creates a drive or at least single directory on your local computer which is 100% supposed to be synchronised with the Google cloud Drive. Any changes to the Cloud or local files are synchronised in either direction.
I've modified this:
Please refer to the simplified diagram above
I have a set of source data. Actually on > 1 disk. For arguments sake it is a SINGLE DATA DISK shown above
I only want to put selective parts of this dataset into Google Drive
Periodically Syncback takes these selections and pushes them to a CLOUD LOCAL physical Disk.
The Insync paid Client, which in testing, unlike the Google free client, has proved 100% reliable, takes those files/directories from the local physical CLOUD DRIVE and synchronises them with Google Drive in the cloud
This Solution means:
a) Google Drive is effectively usually (but not necessarily) a read only repository
b) Because I can't totally trust it!
c) I'm able to custom form a local Google Drive that is a composite from many data sources. Google Drive native forces you to sync a directory only.
d) My Data is still mastered locally
e) But offsite I can always get to any data I need via Google Drive
f) The Syncback mechanism populates my local Google Drive collection and other Cloud providers I use such as SkyDrive(Onedrive) or DropBox. Example, since I am still testing Onedrive my receipts directory is copied to both Google Drive and Onedrive.
g) Offsite if I have need to transport bulk data I can upload it to Google Drive, then download it when I return home. This avoids carrying sensitive information which could be stolen or seized.
g) I am not using Google Drive for Offsite backup. I use Crashplan for Cloud Backup/ Restore
The upgrade was very smooth.
Another Performance Test
I'm doing some Graphics work this week so:
- I populated my local Google Drive with 25GB test files (160 seconds from the NAS)
- I checked that Insync was set to consider my new directory of graphics - yes
- I went for a coffee - 10 minutes
I came back and found Insync uploading at about 40M bits per second. A little slow if you ask me, since even after allowing for the substantial web traffic downloading Vintage computer data, speedtest.net still reported a 79Megabits per second available upload bandwidth.
I wonder who is slowing things down: Insync or Google?
- Insync Plug: The free Google client crashed regularly, had catastrophic syncing problems, and never uploaded quickly. Please see the Blog article. Try Insync.
Whilst I still have reservations about the capabilities of Google Drive I now have a mechanism that allows me to use it reliably and remotely.
The new Drive price plans are more competitive than other Cloud providers. Drive is now again my Cloud Storage provider of choice for Bulk Data.
Google Drive Settings
Dropbox Pricing Plans