Me: I live in Silicon Valley with my wife, child and cat. I have worked at Microsoft since I graduated from College, both in the Macintosh Business Unit on products such as Outlook Express, Entourage, IE, and Virtual PC and in Windows Live on Hotmail, Calendar and People. I am currently a Principal Lead Program Manager on the Windows Live Social Networking team. I basically manage a team of Program Managers responsible for delivering features to support our web and client applications. I've been blogging since 2001 and like to play around with .NET in my spare time working on projects such as dasBlog (the blog that powers this site) and Send to SmugMug (an application for uploading photos to SmugMug). I blog about a number of technology and productivity related topics.
Powered by: newtelligence dasBlog 2.3.9074.18820
Disclaimer The posts on this weblog are provided "AS IS" with no warranties, and confer no rights. The opinions expressed herein are my own personal opinions and do not represent my employer's view in any way.
© Copyright 2010, Omar Shahine
E-mail
What are people using to backup to Amazon S3? I've tried a number of tools and they all fall short for one reason or another.
The basic problem with Amazon S3 is this:
Jungle Disk
Great app, it allows you to map a drive letter to your Amazon S3 account. It has built in backup, however, it does not remove things in the destination that no longer exist in the source. I personally want my backup to be a mirror of my source because I tend to organize/move files around a lot.
Jungle Disk + SyncBackSE
This combo proved to be horrible because SyncBackSE relies on file date/time modifications to detect changes. The end result is that lots of unchanged files get sync' ed again. To make matters worse, when SyncBackSE copies a file to a remote drive, it will upload a temp file and then rename it. Well since Amazon S3 does not support rename, this results in two file uploads by Jungle Disk (and wastes time and bandwidth). SyncBackSE has an option to use file hashes to detect changes, but this is far slower and requires that each file be downloaded to get the hash. No good for remote backup where you are paying for the bandwidth.
This combo proved to be horrible because SyncBackSE relies on file date/time modifications to detect changes. The end result is that lots of unchanged files get sync' ed again. To make matters worse, when SyncBackSE copies a file to a remote drive, it will upload a temp file and then rename it. Well since Amazon S3 does not support rename, this results in two file uploads by Jungle Disk (and wastes time and bandwidth).
SyncBackSE has an option to use file hashes to detect changes, but this is far slower and requires that each file be downloaded to get the hash. No good for remote backup where you are paying for the bandwidth.
Jungle Disk + SyncToy
If you configure SyncToy to "echo" files from your local machine to Amazon using Jungle Drive you get what appears to be perfection. However, I have noticed that renames and deletes on the source are not mirrored on the destination. I suspect this has to do with either a bug in Jungle Disk or some issue with amazon's lack of support for rename. Furthermore, SyncToy does not delete directories on the destination using echo.
If you configure SyncToy to "echo" files from your local machine to Amazon using Jungle Drive you get what appears to be perfection. However, I have noticed that renames and deletes on the source are not mirrored on the destination. I suspect this has to do with either a bug in Jungle Disk or some issue with amazon's lack of support for rename.
Furthermore, SyncToy does not delete directories on the destination using echo.
S3 Backup
I'm currently trying this out, but it lacks any backup scheduling support. I also do not know if it will delete files on the destination that are no longer on the source. I Will find out soon enough I guess. update: looks like this scenerio is possible, but the option to delete files on the destination that are no longer on the source is greyed out in the current beta. hmm.
I'm currently trying this out, but it lacks any backup scheduling support. I also do not know if it will delete files on the destination that are no longer on the source. I Will find out soon enough I guess.
update: looks like this scenerio is possible, but the option to delete files on the destination that are no longer on the source is greyed out in the current beta. hmm.
Anyone else have any tips here?