NAKIVO Backup & Replication Components: Transporter
Sergei Serdyuk, posted on August 12, 2014
The Transporter is an application, which performs all of the data protection and recovery tasks: data read, compression, deduplication, encryption, transfer, write, verification, granular and full VM recovery, and so on.
For the sake of performance, we have designed the Transporter so that it uses the maximum available resources to achieve the top speed of VM backup, replication, and recovery. In other words, it is a server application that should be installed on a dedicated physical or virtual machine, so that it does not interfere with the performance of other applications. This doesn’t mean, though, that the Transporter will always use 100% of the available CPU and RAM: the speed of source datastore, network, and target datastore can be the bottleneck for performance.
Initially, the Transporter (as well as the Director) could be installed only on Windows. Soon after releasing v1 we received a number of requests to support Linux as well. Customers told us the benefits of Linux are that it’s free, requires less maintenance than Windows servers, and is more secure. Since the core of the product is written on (cross-platform) Java, we added support for Linux in a couple of weeks after releasing v1. So now it can be installed on Windows, Linux, or deployed as a pre-configured Virtual Appliance.
To make the initial setup simple and intuitive, an instance of the Transporter is automatically installed and configured with the product installation to enable backup, replication, and recovery out of the box. After installation, it runs as a service, so it’s automatically started if you restart the OS.
A single Transporter can back up, replicate, and recover multiple VMs from multiple vCenters and standalone ESX(i) hosts. In other words, you don’t need to install it on protected VMs or ESXi hosts: it should just be deployed somewhere in your infrastructure where it can access your hosts and backup repositories.
One Transporter can simultaneously process multiple (6) VM disks by default, and you can change this number in the configuration. If your jobs contain more VM disks than the Transporter is set to process simultaneously, the disks will be put in a queue and will be processed once the Transporter frees up.
To achieve scalability and high availability, multiple Transporters can be deployed in a single or multiple sites. If you plan to transfer data over WAN without a VPN connection from your source site to the target site, make sure the source and target ones are added to the product using external IP addresses or using DNS names that can be properly resolved in WAN (so both of them can connect to each other).
If you have more than one Transporter, it is important to determine which one should be used to read data from the source host and which one should be used to write data to the target host. By default, NAKIVO Backup & Replication automatically determines which one should be used by measuring the host proximity based on the ping round trip time:
In the example above, Transporter1 will be selected to read data from the Source ESXi, and Transporter2 will be selected to write data to the Target ESXi. The Transporter selection can also be configured manually during the job creation.
Deploying multiple Transporters also enables network acceleration and AES 256 encryption of traffic between a pair. For example, if VMs are backed up over WAN to an offsite location, the Transporter installed in the source site can compress and encrypt VM data before transferring it over WAN, and the Transporter installed in the Target site can unencrypt the data prior to writing it to the Backup Repository: