The NVA Blog

Back in April this blog mentioned Tru.Vu, a new company started by myself and Tom Chikoore. A few quick updates for those that have been tracking the company.

• The company is now Filtrbox, Inc. We changed the name recently for a variety of reasons, but suffice it to say we are happy with the new name too.
• New site is up, as is the blog. Check out http://filtrbox.com and blog.filtrbox.com for news and info on our progress. You can sign up for access to the preview version when we launch it later this year.
• TechStars is going very well. The program has been very well run and been immensely valuable to us. The program wraps up mid-August with a big demo/pitch day and we are all working our butts off to get ready.
• Another TechStars company, MadKast, has a cool social bookmarking/link sharing service they recently launched in Beta. It’s running on the filtrbox blog, and a whole bunch of other sites. Give it a try or get it for your own blog!

So the NVA blog isn’t necessarily for product reviews but after my experience with Mozy the last two weeks I needed somewhere good to post this. I started using Mozy for online backups a few weeks ago, and low and behold my laptop hard drive bit the dust last Tuesday. In a bad way. The good news is I’m now back up and running with a new hard drive and I eventually did get all of my data back from Mozy. The bad news is that it was very painful and frustrating and their system doesn’t quite work as assumed.

• The way it works is that they break up your fileset into small chunks (encrypted), and spread it out over multiple servers for security. While this sounds like a smart architecture, it took them over 12 hours to rebuild a 23 gb fileset. Waiting 12 hours for an email that says “you can now download your data” while you are in panic stations because you think you just lost everything isn’t reassuring.
• I’m on a 10+mbps broadband connection and regularly get 2-3mbps download speeds. The best I got from Mozy while trying to download said archive was a peak of 700kbps and an average of around 300. This means it took an entire day to download the restore archive, which still failed due to the next issue.
• Their system is flawed in that it removed the restore archive from it’s location the instant it thinks the file transfer is complete. I was using a download manager even, to ensure if the transfer got interrupted I could resume, but at 99% downloaded Mozy thought the transfer was complete but my download manager did not. They deleted the file on the server and I ended up with an incomplete archive that would not mount. This means I was sitting with 22 GB of a 23 GB archive, and had to start over! As in request the restore archive again and wait another 12 hours to try to download it, not sure if I’d have the same problem at the end of the process. At this point I was w/o my data for about 2 days and starting to really lose it.

Eventually I figured out a few things that may be helpful…

1. Don’t request your entire dataset in a restore request. When I broke it up into much smaller chunks it took less time to generate the restore archive and downloading actually worked.
2. Mozy stores file and folder level permissions in addition to the file itself. This is good in general but creates problems when you have to do a clean install of an OS and the UID/GID are not the same. I had to manually reset permissions on most everything I downloaded to get it to work.
3. Their system is best suited as a file/folder level safety net. It’s easy to restore specific files/folders via their desktop client, and this doesn’t have to request a restore archive to be built on the server. Perhaps obsolete to Mac users once Time Machine comes out, but given their price it’s still worth using.
4. They say restore files are available for a week to download. This is only true IF YOU HAVE NOT DOWNLOADED the file yet. They delete the archive as soon as they think you have a copy.
5. As it turns out, if you have a big dataset, requesting a DVD to be Fedex’d to you might be the most painless way to get your data back, but it’s an additional cost.

So in the end, the system does work. I got my data back and was able to use it. I’ll keep using Mozy as a safety net because it works nicely in the background and automatically saves diffs of files I create or modify. But I’ll also begin doing weekly drive-level disk images and save them on an external drive in case a “bare metal restore” is needed. If your system gets stolen, dropped in the lake, etc, using Mozy to get everything back in a jiffy is not the way to go.

Update 8.7.07 - Since posting this I have been contacted by some folks at Mozy who took my complaint seriously and mentioned they were working on improving the restore process of the service. Also, I came accross this article on CNET from Michael Horowitz. He’s not a fan of the service, and highlights things he doesn’t like about it. Some are personal opinion, and some seem to be service limitations all users should be aware of.

I wanted to turn you on to a great post over at Redeye VC about the risks of high valuations and the unintentional impact it can have down the road. Many entrepreneurs think “the higher the better” when it comes to getting their company valued and raising money. Sometimes this really screws things up and forces the company down a path of having to execute the “go big” plan when it might not be the right thing for the business. The more money in, the higher the exit needs to be to satisfy institutional investors down the road. Josh has a great perspective as a seed-stage institutional investor on this topic, and the post has a number of good insights for entrepreneurs, especially if you are looking at a second round and just walked out of the meeting with your partners where you said…”I like firm A better, but firm B thinks we are worth more…” You are perhaps walking your own plank without knowing it.