Omni Loader can seamlessly utilize a number of local machines working in parallel to speed up the migration process.
.svg)
In order to get started, select the orchestrator machine and run Omni Loader in Cluster Mode. Then, run Omni Loader Agent on any additional machine which you wish to be part of the cluster.
You might find yourself in a situation where you need to migrate billions of records and finishing the process on your machine alone would take time you do not have.
Naturally, it would be optimal to vertically scale your machine for better performance (upgrade to a more powerful CPU). If this is not possible, horizontal scaling is the next best solution.
Joining machines into a cluster is effectively an alternative to adding more cores to your machine.
A common scenario people face when migrating or comparing data within clusters globally is latency, especially when migrating an on-premises database to the cloud.
Because the data needs to be movied across the wide area network (Internet or dedicated pipe), latency becomes a problem.
Omni Loader minimizes latency by executing processes parallelly. Workers are simultaneously performing multiple tasks. While awaiting acknowledgement on certain network packets, other network packets are being pushed through the pipe, effectively cancelling out the latency.
While the objective behind Omni Loader’s Cluster Mode is to save you time and run your desired tasks as quickly as possible, it does not complicate the process at all.
It is designed to make the user experience as straightforward as possible while doing all the hard work behind the scenes.
Omni Loader communicates with the agents through a WebSocket connection which allows for ultra-fast communication. It sends commands and receives status updates nearly instantaneously. Each worker has a dedicated connection to the source and/or target database and does a full range of tasks as required.
.png)
.svg)
