Switching from Jungle Disk to SpiderOak

closeHey, just so you know ... this post is now about 11 years and 5 months old. Please keep that in mind as it very well may contain broken links and/or outdated information.

Back in October 2007 I started using Jungle Disk and Amazon S3 (Simple Storage Service) for my online backups. At the time, it was one of the cheapest and more reliable ways of backing up my important documents across multiple platforms (Windows and Linux) to the cloud. Even after Rackspace acquired Jungle Disk in 2008, I continued to use (and be very satisfied with) the service as I was grandfathered into the one-time $20 payment for lifetime use of the client along with the $1/month Plus service for web access and uploading changes only (instead of the entire file, to save time).  In 2010, Jungle Disk released an iOS app that let me access my stored files on my iPhone. But after all of that, support seemed to drop off and the last software update (v3.16) was back in May 2011. Still, everything was working fine and I wasn’t really considering switching, until I heard Steve Gibson echo a similar sentiment about the fall of Jungle Disk on episode #373 of his Security Now! podcast.

I decided to start looking into possible replacements for Jungle Disk, so I went back through episode #349 of Security Now! when Leo and Steve discussed some other available cloud storage solutions and really liked SpiderOak due to its “no knowledge” security policy, which fit with Steve’s mantra of “trust no one” (TNO).

Pricing

My current JD/S3 backup was roughly 75gb in size and costing me about $10/month. SpiderOak gives you 100gb for $10/month, and if you pre-pay a year, you get two months’ free ($100/yr for that 100gb). For some people, the S3 model might be preferred since you only pay for what you use. But in my case, for the same price I was getting another 30gb of storage capacity with SpiderOak. You can also back up any number of devices to your SO account, just like Jungle Disk.

Access

Also like Jungle Disk, SpiderOak offers Windows, Mac, and Linux clients to handle the local backups. You can also access your files through any web browser or supported mobile device (currently Android and iOS) but you cannot back up your mobile devices to SpiderOak.

Setup

After creating my SpiderOak account on the web site, I downloaded and installed the software on my desktop PC, laptop, and Linux server. The SO client makes it easy to quickly select all your Documents, Movies, Music, and Pictures, assuming they are in their normal default locations. I prefer a more granular approach so I used the advanced option to manually select the folders, subfolders, and files I wanted backed up. After getting the backup jobs configured just right, I kicked off the initial upload and let them run … for a few days.

Yes, it took just over 5 whole days to upload ~75gb to my new SpiderOak account. I expected this, but it had been a while since I last did such a huge chunk of uploading at once. In fact, I ended up upgrading my RoadRunner service to Turbo (20mbps down/2mbps up) a few days after starting the upload, effectively doubling my speed (and hopefully cutting off a few extra days of the initial transfer).

What I Like

I’ve been using SpiderOak for about a week now … what’s cool about it?

  • de-duplication: SO can detect duplicate files (say, the same MP3 file) across your backed up devices and then only store a single copy of that file on its servers. This feature allows you to effectively store more data than you are paying for (3 copies of that MP3 file might take up 12mb of data across your three machines, but only 4mb in SpiderOak).
  • historical versions: every time a file changes, SO stores a historical version of that file, allowing you to go “back in time” and restore an older version if necessary. Like de-duplication, SO only stores the changes to the file, not another full copy, which takes up less space.
  • Sync: I’m listing this under “like” even though I’m currently not using it. With SpiderOak, you can create a more-secure-than-Dropbox folder (or set of folders) that synchronize across all your SO devices.
  • ShareRooms: similar to Dropbox’s public folder, or Jungle Disk’s public file shares, you can mark one of your backed up folders as a ShareRoom to generate a unique URL you can share with friends and family. ShareRooms are not indexed by search engines, so they’re not publicly discoverable unless you know the address. Of course, if someone you share the URL with posts it on a forum, which is then indexed by Google, your ShareRoom is now easily findable, so use this feature with caution.

What I Miss

SpiderOak is nice, but I do miss some of Jungle Disk’s features, namely:

  • Windows drive mapping: JD let me map my S3 bucket as a drive in Windows Explorer for easy access to my files like any other drive. With SO, you need to download the files back to your PC through the client software before you can view them.
  • storing only in the cloud: since I had the Windows drive access through Jungle Disk, I could create folders in my S3 bucket and store files just in the cloud that no longer existed on my PC (for example, annual backups). With SO, if you delete a file, or remove it from your backup, it ends up in the Deleted Items folder in your SO account.
  • excluding file extensions/file names at a per-directory level: in Jungle Disk, I could specify exclusion criteria at the folder level. For example, when backing up my video library, I only wanted to back up the metadata (.txt and .jpg files) not all the videos. With SpiderOak, you can only set exclusions at the global level, so if I added .txt to the SO exclusion list, none of my .txt files would get backed up.
  • password-protecting the client: I run SO on my daughter’s laptop but I don’t want her messing around with the GUI and changing settings. If I turn on the password-protect feature of SO, the client won’t run at all until the password is entered. To get around this, I had to schedule SO with –batchmode via Task Manager in Windows so it runs silently in the background and she’s not even aware of it. I guess this is slightly better than JD since now there’s not even an icon in the tray for her to notice.
  • detailed reporting: with JD, I could set up backup reports that were e-mailed to me on a weekly basis so I could review what files were backed up. This was especially helpful with regards to my daughter’s laptop so I could confirm her files were being backed up every day. SO has no such reporting facility and in fact every time you open the client, the Log tab starts empty. I found some log files in the %APPDATA%\SpiderOak directory under Windows, but they’re not exactly easy to review. Going through the View tab in the SO client trying to determine what files were backed up yesterday or the day before is a bit tedious, if not almost impossible.

Problems/Bugs

The biggest problem I am currently having with SpiderOak is on my Linux server: every time I open/run the SpiderOak client, it does a re-scan (which I understand is as expected behavior) and then re-adds every file in the backup to the queue, creating a new historical version. This results in a queue of about 46,000 items to upload which takes a long time to complete (about 3 hours). This isn’t happening on my Windows devices, so it has something to do with how the Linux client is determining if a file has changed. I’ve posted about this in the SpiderOak forums and am awaiting a reply from support staff.

A minor bug is that the GUI doesn’t remember its window sizing, so every time you open the application it reverts to the default size. This wouldn’t be a huge deal except none of the columns are resizable like they should be … you have to resize the entire window in order to see more in each column.

I’ve already mentioned there is a lack of user-readable logging across SO sessions, but I’d also like a better way to determine which files in my backup have been de-duplicated. Right now the Stats area of the Status tab shows your total backup size before de-duplication, but it’d be nice to know what files in the backup are duplicated across my devices.

Finally, I have issues with the scheduling. There is no SpiderOak service and the scan/backup only runs when the client application is running. This means if your machine is on, but you’re not logged in, the backup won’t run. This can cause problems on a multi-user PC. For example, on my daughter’s laptop I logged in as myself when I installed and configured SpiderOak. If she’s logged in, the backup doesn’t run. To get around this, I had to schedule the SO backup via the command-line and the Windows Task Scheduler. It’s not an ideal method, but it seems to be working so far. This is just another extra hoop to jump through.

Conclusion

It’s only been a week (and most of that was just uploading my files), but so far I like SpiderOak. Despite the few snags I ran into, the service works and the price is definitely right for me. Once I figure out that problem with my Linux backups everything else should be smooth sailing. I’ll be sure to comment on this post over the next few weeks as I use it more, but I expect to cancel my JD/S3 account by the end of the year.

7 Comments

  1. Too little, too late? Jungle Disk apologized for its lack of customer focus and promised exciting changes coming in 2013.

    I’ve been using SpiderOak for about three months now and am satisfied with it. I’m still adjusting to some of the things I listed in the post above (and SO support is still trying to figure out my problem with the Linux client re-uploading a new version of every file every time the backup runs), but at this point going back to Jungle Disk just isn’t in the cards for me.

  2. I think SpiderOak could not show you what files are duplicated across devices because the deduplication is not file-level dedup but block-level dedup. Thinking again, I think they could show you duplicated files, but it is not important (as far as the backup process is concerned) because when hashing/building the data blocks only one copy would be uploaded for a given account (dedup is per-account, not per-device).

    We too use SpiderOak but only on Linux servers and using the command line in batch mode (from a cron job we call a Bash script which do some logging, some checks and then call SpiderOak). It works most of the time, but sometimes it just hangs there doing nothing (well, it is doing something, according to the logs in the .SpiderOak directory it is waiting for… something). What we have done is detect a stalled SpiderOak process (by counting the “but not b/c mutex locked” in the logs) and kill it before calling it again.

    BTW, file/dir exclusion from the command line version is a pain to set, as there is no obvious way to use wildcards (only specific full paths).

    But it works. We usually test the restore process once a month just to be sure everything is working.

    We tested several backup options but SpiderOak was the only one which worked from the command line, is multiplatform, does block-level deduplication and the data is really private.

    Regards,
    MV

  3. v5.1.3 of the SpiderOak client seems to have finally fixed the problem where every time the SO backup ran on my Linux server it was uploading a new version of every file. My nightly backup runs a lot faster now since only changed files are being uploaded like they should be.

  4. One other caveat about backing up large files with SpiderOak: the client needs to make a copy of the file in order to encrypt it before uploading. As a result, you cannot back up a file larger than the available free disk space you have on the machine running the SO client (see this article for details).

    I found this out the hard way when I ran out of disk space on my Linux server where I run the SpiderOak backups. I had added two laptop images (Windows backups) to the NAS, each of which had a 100+gb VHDX image file. When the SO client tried to copy that large file to the Linux box to encrypt, it filled up the disk and caused all sorts of problems.

  5. I just learned today that SpiderOakONE Backup is in “maintenance mode.” I realized that there hadn’t been a new version of the client since the last 7.5.0 release back in February 2019 (for Linux) and 7.5.1 in September 2019 (for Windows). I’ve asked for more details about what this means. It concerns me that maybe I need to start looking for a new zero-knowledge cloud backup solution. My current SpiderOak account has 5.8TB in it (without de-duplication).

    • Per my inquiry to SpiderOak support, maintenance mode “… means that there will be no updates with improvements or new options for now, but it continues working and we continue providing support. Your data is safe and stored.”

      So I guess if it’s not broke …

Leave a Reply

Your email address will not be published. Required fields are marked *