Ashar Baig's blog

Five reasons why MSPs currently offering hosted services should be looking at this growing Cloud Backup market to move into this market.

1. On-demand Business Model – The cloud-based on-demand business model is here now – the new economic reality is that the old asset-based business model is giving way to OPEX-based approach
2. Magnification of MSPs size and scope – Cloud Computing eliminates the geographic boundaries for MSPs to target and attain potential customers and allows them to expand beyond their previously perceived geography and scope, magnifying their business expertise, thus allowing MSPs to expand their business
3. Operational Efficiencies – MSPs providing cloud backup and recovery have expertise that allow them to provide services more efficiently than in-house IT departments. Due to the economic downturn, customers are smarter about their IT strategy and are more open to entertaining IT as a Service (ITaaS) and take advantage of the cost savings, due to better operational efficiencies, offered by the MSPs.  
4. Forward Thinking MSPs – Cloud Computing elevates how MSPs think about their business. This new business model enables the MSPs to acquire new cloud-based backup and recovery skills with little effort.
5. Low or Few Barriers to Entry – Cloud Computing lowers the overall cost of entry into the market for new MSPs.

The above reasons present business opportunities for MSPs. Cloud backup enables MSPs to grow their business by winning larger customers that are farther away. It gets MSPs in the door so that you can sell more of your expanded services resulting in incremental revenue.

Cloud business model is more acceptable today because the customer problem is being exacerbated by data compliance and disaster requirements. Data protection has not kept up with the backup and recovery resulting in an increased need for more resources and human capital, which makes cloud backup and restore a much more cost effective solution. Cloud is more efficient – when the data grows, the ability to manage that data does not scale linearly with the effort to manage the data.

Enterprises and vendors alike often focus so much on data backup that sometimes they forget about the reason that they backup the data. Customer's focus should be on data Recovery not data backup.

All vendor solutions in the marketplace backup customer data but it requires real data stewardship to ensure that the data can be restored when needed. Over our 24-year history, Asigra has developed best practices around data stewardship to ensure data restorability if the customer looses a file, disk, machine or the entire facility.

The data has to be conditioned constantly to ensure restorability. The following factors can cause data corruption:

1. Disc malfunction
2. Disc controller malfunction
3. Bad sectors on the disc
4. Filesystem corruption

Access to metadata is not sufficient because bad sector on a disc can render metadata unreadable.

Following data integrity and consistency check functionality is embedded in Asigra software to ensure, data restorability:

1. Ensuring data consistency – this process ensures that all the data components have been collected sequentially by the DS-Client (the data collector at the enterprise customer’s premises) before sending the data to the DS-System.
2. Ensuring all data has arrived offsite before storage – Asigra’s DS-System (the online data repository) writes all the data being backed up offsite to a temporary location, checks and ensures that all the data has arrived before storing it.
3. Restore validation – this is an actual restore simulation that conducts an actual data restore to a temporary location to ensure data restorability. Think of it as the data restore dry run to prepare for the actual disaster.
4. Autonomic healing – this automated process runs on the DS-System in the background, scans the entire storage to ensure data integrity. Since the data at the DS-System is encrypted, the “Autonomic Healing“ process checks links between the data blocks, compares digital signatures between different components for inconsistencies. When corrupted data is uncovered, it is marked as corrupted and a notification is sent to the DS-Client to resend the portion of that data that was marked corrupted. This ensures that the data is always recoverable in case of a disaster.
5. Backing up the DS-Client database to the DS-System – this ensures that if the DS-Client is lost it can easily be rebuilt with the appropriate backup structure.

When you're shopping for a backup solution, please inquire from your vendor to ensure that the functionality they provide will restore your data, not just during a Disaster Recovery (DR) drill but in the event of an actual disaster (accidently deleted file, damaged hard drive, machine loss or lost site).

Agents are usually presented by the vendors as very light consumers of hardware resources (typically 2% over a 24 hour period). This resource utilization is accurate when the agents are sleeping but at peak times the resource utilization of agents can increase to 20-25% of the total system hardware resources. Additionally, a single agent adds as much as 16% server overhead to each application.

In physical server environments, there are typically 1-2 agents per physical machine. The same number of agents are required per virtual machine in a virtualized environment. Since a single physical server can be transformed into tens of virtual servers, the result is that there are lots of agents per virtual server host, taking up most, if not all, of the system resources. Moreover, at peak times the system hardware resources are oversubscribed, killing the I/O, rendering the server useless for backing up your critical data seriously impacting your RTO and RPO.

Consequently, backups will need to be scheduled serially extending the backup time windows. The bottom line is agents make data backup and restores difficult.

A virtual environment is the perfect use case for the need of agentless backup and restore – it supports a multitude of applications and platforms without the need of an agent.

Users trying to do backup and recovery should be able to install a backup and recovery solution on physical or virtual machines and have the flexibility to define a physical or virtual machine as a backup target. MSPs should be able to pre-install the client backup software on a virtual machine or a physical machine and ship it to their customers.

Additionally, users should have the flexibility to backup data at the disk level (backing up the entire virtual machine image) or at the application or file level. The ability to backup data at the application or file level enables granularity and is extremely useful when restoring data, saving wide-area bandwidth and storage costs. 

References:

Image Source: Flickr User: jimmyroq