This first webinar was not recorded because of technical problems. Future ones will be recorded and posted on our website.

THanks
Frank

The other thing that client’s like the idea of but I have not figured out how to deliver at a price they are willing to pay is that in the event their machine dies running up a VM “in the cloud” with all their stuff until a replacement is ready.

Lots of good discussion on this topic! I like it. No doubt Acronis is good stuff, and it has been around a long time. But, I think you have your prices or products wrong.

Acronis True Image is the 3 for $79 product you’re talking about. It is supposed to be used for home use, and yes you can get three of them for $79. They have a very limited feature set, and are no good for business use like most of our Service Providers need.

If you want to compare apples to apples, compare Acronis Backup for Workstations, which costs $75/ each or the Advanced version for $99/ each.

Further comparing Apples ot Apples, the Acronis products in the Server catagory (where most of our Service Providers need a solution) Acronis’ products cost $499, $853, $1,399, and $1,799 – more in line with what we propose to compete with.

So, while I think you have a valid point insofar as consumer-grade image backup is concerned, you missed the (price) point on the products that really matter in the Business world.

Acronis has a long-standing reputation, and their products are reliable. RBS, being far less known in the industry than Acronis, will have their hands full charging prices higher than Acronis. IMO.

Most of our clients reside in our city. Considering that this service is sold to our clients an an ancillary service and that we are in fact the IT Department for most of our clients it stands to reason that we’ll be the ones helping out if/when things go horribly wrong. We’ve been using Acronis to create images of our client PCs in our shop and it has been working very well. We’re also accustomed to dealing with VMs. When Acronis gave us the ability to perform a Universal Restore, we got very excited. At last, we have the ability to restore a client’s backup to either alternate hardware or a VM so we can have hardware on standby for this purpose and restore their data locally to alternate hardware in perhaps 4 hours. This is a service you can charge for AND feel good about.

If, like us, you’re going to get the call when there’s a fire, theft, lightening strick or whatever, then I suggest that you’re going to want to consider the costs and challenges of restoring from System State backups on a Microsoft SBS Server. A one-time fee of $500.00 is nothing compared to the 2-3 days that it could take you to complete this disaster recovery task. Add to that the fact that your client will be looking over your shoulder and asking why it should take this long to recover when they’ve been paying you for a backup solution all along.

I think ideally, the best solution would involve a Windows Image Backup to an alternate HDD and then having a bit-level backup of that Image synchronized to the RBS Server. This way only data that was changed since the previous backup would be transferred and the backup provider would have the ability to restore the entire Image to a VM in the event of a disaster.

I feel that, in regards to required storage space, we may not be required to keep multiple copies of a server image. Multiple copies of data is good but a single image should be all that is required for a disaster recovery — especially if we can keep 30 days of data files alongside it.

Cheers,
Jason

I am really scratching my head here…as to why anybody would want this capability (or want to pay for it). All Windows machines since Vista come with a built-in volume-based backup utility supported by Microsoft.

BMR (image) programs pretty much REQUIRE that you backup the entire volume – and thereby effectively need an amount of network storage which is greater than the size of the real drive.

If you needed to do a full BMR restore, you would need a Windows distribution volume to boot from (or a boot disk if you created one) but from that point forward, the system and all settings plus all apps, etc will be restored using Windows Backup. This capability is included in the base Windows OS at no additional cost.

To even think about moving 100′s of GB of data over the Internet (contents of a full image backup) is extremely naive. And, as Rob points out, the backups are large, (even the “incrementals”) and need to be shuttled to the server and the restore needs to be shuttled to the client, so really, what’s the point or where is the value add? (If you manage all of the client’s IT needs, then just rotate out a Windows backup disk every time you do a preventive maintenance call or onsite visit. Right?

Also, from a technical perspective…in order to create a stable, bootable, point-in-time image (on a live system), there is little choice other than to depend on VSS (Volume Shadowing) which itself can be very fickle and require substantial VSS shadow storage be available.

Any software which attempts to do a block-by-block disk backup without either shutting down the OS or invoking VSS…will create an image which will by definition have some “corruption”. I’m thinking about Exchange message stores and other mission-critical databases which cannot be “fixed” with simple chkdsk commands (as explained in some of the previous BMR documentation I read from the beta) Perhaps the close-to-finished RBS BMR offering does now use VSS?

One of the major downsides to using the Windows built-in Backup capability for BMR is that it is an unmanaged program. By that I mean if anything goes wrong…and you are not regularly looking at the event logs…who would know? I mean, its not like you get a nice email each morning that tell your if your windows image backup (full or dirty-block incremental) completed without issues.

Providing a management interface that allowed you to treat the Windows backup image backup … as another Rbackup job … now that would be interesting and value added. In fact, I believe that would adequately suffice for “most” folks image backup needs. (perhaps position as a plugin that sends across a session file with some Windows backup stats …similar to how the SQL plugin uses SQL to do the actual backup and reports out SQL error codes.) That would perhaps command a 1-time $499 fee (I would strongly consider it) since it would let me remotely manage my customer’s full backups. Remember, the critical data is offsite using Rbackup anyway!

Now, having said all of that, if the backup provider offered “a virtualized system restore” i.e., if the client had a crash, then their latest .vmdk or .vhd file would be “booted” into a virtual machine…that could be interesting. However, we are now moving from a product and pricepoint which can attract a large number of prospects…to a much smaller set of target prospects that have short RTO’s and enough money to buy what they need. Probably not the typical SMB marketplace. And, again, as Rob points out, any failures would be an immediate 24×7 crisis situation to deal with.

fwiw, I would also definitely pay for a plugin that swapped out the Rbclient backup.mdb file for a Microsoft SQL/Express high-performance database. Similar to the SQL upgrade available for the RBS Server. The increased performance would allow us to scale to larger backup sets for higher-end clients with 100,000+ files to backup (and synchronize). I would say a one-time cost of somewhere between $599 and $999 for a high-performance client SQL-based would be in the right ballpark.

regards,
Mitch

How about alternative pricing, like $15/month per user?

Breaking data files up into blocks and performing block-level deduplication offers the possibility of increasing the hit rate i.e., being able to store a reference counter to a hashed key vs transmitting the entire data block each time.

However, in both cases, if you mandate that insecure data — even in chunks — cannot — at any time — exist in unencrypted form off the customer’s machine, then I believe you reduce the benefits of block dedup to pure random chance. Which is I believe the crux of Rob’s argument.

That being the case, any benefits associated to pure chance “hits” should probably be handled by dedup on the destination file system using “background” processing thread.

If however you remove the requirement that data be encrypted on the customer machine…and permit an SSL link to carry these common data blocks over the wire (which can be subsequently hashed and encrypted) to the data center, I can see how block-level deduplication could yield substantial data storage reductions.

These benefits would however come at the cost of increased overall complexity and processor overhead in the data center. I don’t know how big the data center would have to be for the savings to outweigh the back end infrastructure costs…if it ever did.

So, if Rob’s large prospect permits transmittal of data over SSL links and subsequent processing (encryption) server-side, then the “competitor” could have an economic advantage as the infrastructure scales…since in theory the larger the “base” of reference hashed blocks is, the more efficient the service would become.

I’ll rephrase your question. How do “legitimate news sites” support their sites? Through ads, of course, properly placed in places for ads, and clearly defined as ads. Non-legitimate sites support themselves through sleazy tactics like I described in my article.

The Wallstreet Journal would never refuse to publish legitimate news about a company because that company didn’t buy an ad, or because an editor had a personal issue with someone at the company.

The blog I am referring to is certainly not influential or important to my industry. Those of us in the industry (and even their advertisers) know this about them. It’s just an example to illustrate my point that the general public is being fooled by blogs that masquerade as authoritative, but are really simply complicated self-serving ads pushing for those all important first-page listings on the search engines so they can sell more ads.

I could have picked Wikipedia as an example. What a joke! Many laypeople use it as the go-to authority for most questions. Yet behind the scenes there are a lot of self-serving “editors” who are often paid by companies to alter pages in their favor, or who just want Wiki points, or whatever people like that get off on. Some pages are authoritative and some are not.

Here’s an example from Wikipedia. Look up Online Backup. In the History section you will find a paragraph about Exabyte corporation. Exabyte has NOTHING to do with Online Backup – nothing at all. They are a disk-to-disk company and always have been. They do not belong on that page.

If you check out the history of the page you will see that the Exabyte reference replaced a reference to myself. I’m the guy who actually founded the online backup business and really does belong in the History section.

I don’t care whether my name is on Wikipedia or not. That’s not the point. The point is that the general public mistake sites like Wikipedia and the other blog I ranted about as authoritative because they print what looks like news, and I think that is simply shameful.