RedHat a smart move: buy Gluster
On October 4, 2011, Red Hat announced that it had entered into an agreement to acquire Gluster, an open source storage cluster software system headquartered in Sunnyvale, California, and developed in Bangalore, India.
Gluster technology provides capabilities that make it a good fit for Red Hat’s cloud computing strategy. Red Hat is actively engaged in developing cloud solutions, as evidenced by its
IaaS : CloudForms
PaaS : OpenShift
Gluster technology is designed to handle large scale-out data environments, storing unstructured data. As open source it is designed to run on commodity hardware.
An important feature of Gluster is that being a purely software solution, capable of running on standard hardware, it differentiates it from many competing products, which require the purchase of specialization, specific hardware vendors. In addition, it offers important advantages to the customer: reduced costs, greater flexibility of implementation and reduced vendor lock-in.
As software, Gluster has the advantage of being able to scale in environments ranging from collections of on-premises storage servers to clusters of virtual machines built on Amazon Web Services.
Gluster is distributed across multiple systems and aggregates total storage into a single namespace. A Gluster cluster exposes this namespace as a point of NFS or CIFS that contains all the files in the cluster. The advantage of this model is that the underlying repository becomes fully virtualized, and can also be deployed through public or private clouds.
Gluster overcomes the downside of distributed file systems (degraded performance and scalability due to metadata management) by using a sophisticated algorithm that eliminates the need for metadata servers.
With Red Hat, Gluster’s support for the community will benefit you in the long run.
Starting with a base in government research institutions, Gluster’s customers have grown to include companies such as Deutsche Bank, Samsung, Autodesk, BAE Systems, Barnes & Noble, Dana Farber Cancer Institute, and Pandora.
As described in Red Hat’s press release and blog, Brian Stevens, CTO, “Gluster shines in environments that need to manage large amounts of unstructured data distributed across a heterogeneous, scale-out environment, such as data scattered on-premises and public clouds, or multiple facilities. Examples are multimedia, geophysical/astronomical data, financial analysis/modeling, medical data, social media data streams, etc.”
As VMengine, we started evaluating storage cluster realities back in 2007 at SNIA in Rome with Isilon and LeftHand. Because at that time, in figuring out how to design a cloud hosting platform, we came up against the harsh reality of classic storage systems and their difficulties in terms of scalability, size, and access bandwidth. The first to be acquired was LeftHand from HP. Isilon was acquired by EMC in 2010 .
When NASA released its cloud computing platform, we noticed that it used a filesystem distributed on clustered storage
The world of server virtualization concentrates virtual machines into a few powerful devices, but the growth in CPU power and RAM speed does not follow a corresponding growth in the speed of data storage systems, and the growth in the amount of data to be stored makes classic storage systems expensive.
As VMengine, we have been testing and using the Gluster platform for some time and included in our projects in Public Cloud environments.
In the complex world of storage clustering systems, it is necessary to pay a great deal of attention to the logic of replicating the data distributed between the nodes of the cluster.
Let’s not forget what happened to Amazon due to human error related to these reasons.
More insights into the world of storage, http://blog.vmengine.net/2011/08/03/cloud-storage-we-need-more/
Last update: https://www.websiteplanet.com/
