Product review: Dell AppAssure 5.4.2

Some time ago, I heard some people talking about Dell AppAssure. The company was founded in 2006 and has been acquired by Dell in February 2012. I heard of it before, but never took the time to take a closer look at it.

I find it very important as a backupconsultant to know your competitors strengths and weaknesses and how they adapt their solution to a customer environment. Basically I want to understand why it’s better than another product and the other way around.

As we have our national holiday on monday, we can enjoy a long weekend. I dedicated some time this weekend to investigate the backupsolution of Dell and write something about it. As always I start-off with some reading to understand the concepts, caveats and features off the product. If it’s worth my time, I try to install everything in trial mode to test it out in my home lab environment.

I think it’s important to mention that this blogpost contains first impressions, thoughts and ideas written down during the initial product review.

This review is written by using a personal labo containing:

  • a AppAssure Core Server version 5.4.2 running on Windows 2012 (2 CPU / 4GB RAM);
  • an Active Directory Domain Controller running on Windows 2012 (1CPU / 2GB RAM);
  • an Exchange 2013 running on Windows 2012 (2 CPU / 4GB RAM);
  • a Windows 8.1 computer running VMware Workstation 10.

The following test scenario’s were defined prior to installation:

  • Solely Microsoft based platforms are triggered.
  • Software deployment by using the console: I deployed one agent by performing a pull install and adding it in the console. As a second test, I pushed the installation to a server. During the operations the software was being installed and sequentially a protect operation was executed. This gave time-out and it was required to manually add the server in the protected machines sections. Once done, the protect operation could be restarted.
  • Cross-firewall (ports) backups: checked, but due to infrastructural limitations this was not tested.
  • Pre- and post scripting: OK;
  • Single-file restore: A recovery point of the Active Directory Domain Controller is mounted and a file is restored. The logfile was readible and editable.
  • Single-mail restore: An e-mail is restored by using the Mailbox Restore application. This worked very smoothly.
  • Virtual Standby conversion (export operation): An export of the Exchange server to a WorkStation Virtual Machine is executed and validated.The operation was scheduled as a post-process job after the backup. Once the VM is imported and started, the machine started accepting IO again. 
  • a bare metal restore to similar hardware (VM > VM);
  • Daily basis reporting (SMTP);

First of all, I was pretty amazed with the way documentation was structured! From a technical point-of-view I do require some more details and background info about supported setups, for example: can you split SharePoint Application Server and databases or do they need to be on the same server? It’s a bit too high-level for me at this moment. I miss a search functionality and there is only one thing that really worked on my nerves.. Why don’t put the compatibility charts in the product-specific (in my case AppAssure 5.4.2) documentation? Or at least, create a link to compatibility matrix or something. After some Google searches, I found this link which lists the compatibilty of different systems and software components. And to be honest, I’m not entirely sure this is location that you need to check before deployment. I notice some references to SharePoint (DocRetriever) in the features section, but no SharePoint records (none at all) are listed in the compatibility matrix at all. Above all, it’s very limited! From my point of view, a decent compatibility / documentation page needs to be available, listing all information in a structured – and searchable way. If this is not the case, I tend not to use the product.

The ideas and backup approach are entirely different than other backup products. A backup solution nowadays is in most cases client-server-based (except when using integrations such as VMware API for Data Protection which requires no agent to be installed in a virtual machine for backup purposes). This approach leads to a central location (which is called the AppAssure Core) who instructs backup operations on clients by communicating with a little piece of software (the AppAsure Agent) on the client.


Regular backup products use the archive bit or the Microsoft Journal Log to identify data that is altered since the last backup and requires protection. Dell AppAssure on the other hand uses the VSS components on a server to create VSS shadows (or SnapShots) and transfer these to a backup repository. Other products using this approach are ShadowProtect and MirrorCloud (website in Dutch). This different approach allows incremental forever backups and space-efficient – non deduplicated – data transfer, but I have some doubts about the stability as the software uses VSS quiescing to create a consistent copy of the volume. On servers under a high-load creating and releasing VSS snapshots can lead to some unpredictable circumstances, such as performance degradation on the production servers and/or recurrent backup failures. I encountered similar problems with a rather large and heavily occupied virtual machine which had all data in VMDKs stored. Eventually I continued with backing-up the system drive with VMware integration and protect the rest of the data with an agent. A workaround can be to distribute your data over different disks, limiting the IO impact a disk needs to handle during a VSS-based backup operation, but that’s an entirely different story.


The software uses a volume-level data selection. This means no data can be selected within a subdirectory and the whole system will be protected even though you only need one specific directory. By using this concept, you have less flexibility in selecting and storing data.

Data restore is performed by mouning the recovery point and using a simple browse functionality to restore data back to the original location. Dell provides a function within Microsoft Explorer to restore the data with the correct permissions (“Dell Paste with Permissions”).


Granular restores are performed by using a tool that can read through the application database file (just as Veeam it does with U-AIR). This is called Universal Recovery – Object Level Recovery.


For data retention, we don’t talk about days or weeks or “cycles”. Data is retained by using recovery points (some kind of “cycles”).


Data backup requires a backup-to-disk functionality. The backup repository supports compression and deduplication to store data in a space efficient way. Backup-to-tape is not supported, in contrary we talk about archive-to-tape (like with Veeam). When I take a closer look at it, it’s definetily not a method I’m willing to use. I do think they need to put some work in a decent integration to support archive-to-tape! Today, I have the idea this feature is solely available to check a checkbox in the product requirements section at an offer.


The software itself supports several cloud providers:

  • Microsoft Azure;
  • Amazon S3;
  • OpenStack;
  • Rackspace Cloud Backup Storage;
  • Additionally data can be replicated to another Dell AppAssure platform within a routable LAN segment.

Furthermore, the software provides a Virtual Standby for selected servers. This feature performs backup data conversion towards a pre-defined data format. This allows swift recovery by simply booting a virtual machine. Today VMware ESXi, Microsoft Hyper-V, VMware WorkStation and Oracle VirtualBox are supported data formats.


The software supports dissimilar bare-metal restores. This means in case of system down due to hardware error, the server can be restored to a spare server! Simply by booting a boot CD connecting to the Dell AppAssure Core server. The process is similar to the Simplified Disaster Recovery Process of Backup Exec.


No backup policies… The software allows configuration on a global-level and when rolling-out new servers, you can choose to use the global configuration (for example for retention policy or scheduling). In case, you want to use server-specific or volume-specific approach, the configuration will be time-consuming as each server needs to be configured independently.

I notice the software is a SILO product. It does what it is designed for, but nothing more or nothing less. I expected some integrations with Dell EqualLogic or Dell Compellent but I remained disappointed.

Additionally the software does not support pre- and post scripting of backupjobs. If you want to implement this functionalities, the scripts needs to be copied under a specific folder. This results in not providing a single-pane-of-glass. These functionalities should be a part of the basic features of every backup solution.


  • The backup approach allows close Recovery Point Objectives (RPO) and Recovery Time Objectives (RTO);
  • A good solution for small- to mid-market companies;
  • Some nice features are available, such as bare-metal recovery, SQL verification check and virtual standby.
  • Software that’s is able to read the Exchange database to efficiently restore mail items;
  • Backup results are logged in EventViewer which can be collected with SCOM or other monitoring solution.


  • Limited support matrix – unclear/ vague compatibility charts;
  • No support for VMware VADP;
  • No dynamic inclusion of freshly added volumes, these require to be manually added for protection;
  • Use of shadow copies can have some serious downsides and need to be used with caution;
  • No fully archive to tape support;
  • No LAN-free backups;
  • No integrations with EqualLogic or Compellent;
  • The reports are not customizeable;
  • Error messages are a bit unclear;
  • No “real mail or filesystem” archive functionalities.

If they would ask me where can you fit this product, I would say small- to mid-market enterprises that are standardized on Microsoft platforms and applications. I would not use it if my customer states he needs a long term retention on a removable medium.

Thanks for reading!


Leave a Reply