Going undercover with agentless backup
Gabriel Gambill gives his view on the latest phenomenon being talked about in disaster recovery and business continuity offerings
Most channel people probably know that agents are the small applications installed on a server to perform a particular function.
For backup, the agent is installed on a host server the system administrator wants to back up. Agentless backup is, as its name suggests, backup without the use of such an agent.
To distinguish themselves from their rivals, several backup and recovery vendors claim to provide agentless backup. In many instances, however, these vendors inject an agent at the beginning of the process and remove it before the backup finishes in order to achieve application consistency.
It's up to the channel to distinguish between the different offerings and help customers make a decision. Strictly speaking, vendors aren't providing agentless backup because they are still using an agent in parts of the process.
While agent-based solutions are effective and proven, there are a number of reasons why the channel can move its customers away from agent-based backup.
Some agent-based solutions can be weak from a security point of view and high cost. Adding agents increases the amount of software organisations have to manage.
Agents can also interfere with the processing power of core applications when they are installed.
Even in a normal-sized environment, agent management can be a complex process as IT administrators deal with different agents on different operating systems.
The growing number of software packages running on a host server could prove to be a further complication.
By removing the agent, backup solutions can bypass the process of agent management, while ensuing a faultless backup.
Agentless backup doesn't require a reboot of your system during the installation process. By having only one piece of software to manage, it can work faster and more efficiently in comparison with agent-based backup.
Agentless backup can upgrade itself. Often hundreds or even thousands of systems need to be backed up, so the process becomes less time consuming for IT managers.
Some organisations cannot inject agents due to regulatory requirements. Where data handling cannot be modified from beginning to end, an agentless solution allows for compliance during backup.
Agents can also interfere with the processing power of core applications of the machines on which they are installed. With agentless backup, remote sites can easily implement and maintain data backup programmes.
Going agentless eliminates the costs associated with agent-based backup. The channel can help its partners save money on maintenance and operating costs. Agentless backup may be the holy grail for disaster recovery and continuity vendors, but few have showed signs of achieving this so far.
Faced with a variety of ways of backing up critical data, the channel could educate customers on a number of factors they should consider including cost, simplicity and consistency. Businesses can make an informed decision based on their management, CPU overhead and compliance needs.
Truly agentless backup is not only easy to use but also requires no injection or removal of an agent during the process at all.
Gabriel Gambill is EMEA senior systems engineer at Quorum