Virtualization is far more than a buzzword. When done properly, it provides the means to get more done for less. Grantek has the experience and skillset to implement virtualization and hyper-virtualization properly. We can analyze the server workloads and help our customers determine the exact level of consolidation possible. The average physical server is under 20% utilized on average – leaving a huge opportunity for consolidation.  Virtual servers configured on one physical server—called a virtual host–take advantage of the idle time to do more with less. A virtual environment also greatly shortens new server provisioning time by bypassing hardware procurement, sometimes a factor critical to project success with tight timelines.  Because of our broad experience across industries, we have the expertise to identify how to P2V (Physical to Virtual) or lifecycle systems onto a robust virtual platform.

Virtualization brings advantages to your operations such as reducing the cost of purchasing and maintaining hardware, reducing the cost of power as the number of physical servers is reduced, and reducing the physical space needed to house the servers. Virtualization can also decrease disaster recovery time, increase scalability, and increase availability. Among the benefits of virtualization, disaster recovery ranks at the top along with high availability and decreased deployment time.

Grantek  is a VMWare certified partner and we also work with other hypervisors such as Citrix and Microsoft’s Hyper-V. On the public side, Grantek is capable of deploying SaaS and IaaS solutions in AWS, Azure, Rackspace and the Google Cloud Platform to name a few.  To determine how to properly implement the target platform, we also perform a network assessment (Link to Network Assessment] as part of our site assessment.  Virtualization requires proper design to fit into your system and provide the high availability required for 24×7 manufacturing facilities.

Grantek can help consolidate your older or end-of-life hardware using virtualization, and at the same time provide more comprehensive backup capability. Virtualization offers an opportunity to deduplicate hardware sprawl and achieve efficiencies in the server room.

In a manufacturing facility, typically any device at level 0 or level 1 of the ISA-95 model (sensing and control) stay physical today. Anything above the Programmable Automation Controller (PAC), however, is a candidate for virtualization.  With some generalized rules, Grantek will apply an analysis of your specific needs and business drivers to the recommendations made for server class systems to provide the best balance between consolidation within and across sites, and manufacturing system availability through outages.

All major virtualization vendors offer high availability, which implements a redundant configuration that allows applications running on a virtual machine that experiences a failure to start immediately on the redundant server, increasing uptime and enhancing reliability. This offers more flexibility than traditional hardware mirroring servers, and also eliminates the need for multiple software licenses in some cases, because the software runs on one server or the other, never on both simultaneously, allowing one license to be used for two devices.

High Availability and Fault Tolerance are two availability-increasing technologies which must be evaluated by the underlying applications sensitivity to downtime.  In most instances, problems arise due to an application crash, not the physical failure of the hardware. In a mirrored server configuration or FT implementation, if one application goes down due to a software crash, the mirrored server will likely suffer the same condition.  In other cases where application uptime is paramount a hardware mirror or FT implementation may be required to meet the desired SLA.

Because we have extensive experience designing and supporting manufacturing production environments, Grantek is adept at understanding the level of outage each server can sustain and design a virtual solution accordingly. We utilize store and forward configurations at the PLC/PAC level, so that if a VM receiving the PLC data goes down, the data is stored at the PLC/PAC level and forwarded to the backup VM when it comes online, ensuring no data is lost.  An understanding of the transport layer is paramount to this design; a cellular WAN connection would most likely provide a less robust transport than two systems connected to a current-generation enterprise redundant local network.

A common server setup for virtualization is the 3-2-1 configuration, consisting of 3 servers, 2 switches, and 1 SAN with redundant controllers that is shared by the 3 servers. This eliminates the need to sync the data on the servers because the data from both is going to a common storage device. This configuration traditionally requires lots of cabling, though it does virtualize the compute functionality.

In a hyper-virtualized environment, storage is virtualized in addition to compute via software-defined storage (SDS), which drastically decreases the external network load and is frequently paired with network virtualization. This further increases reliability, reduces maintenance, and increases processing speed, while also reducing network capacity needed. For example, though the typical 3-2-1 virtualization configuration could reduce the amount of equipment needed for 10 2U servers from approximately 20 rack units (RU or U) down to 9U, a hyper convergent setup reduces physical space needed down to 2-4U, and it also serves to provide robust redundancy for all storage, network, and compute components.

Moving an additional step forward from virtualization is containerization like Docker. Where virtualization offers the aforementioned benefits over physical server hardware, it still has drawbacks, the consolidation of a VM stops at the virtual hardware layer.  Each OS is independent and must be maintained and patched separately. Containerization offers the same benefits of virtualization plus abstraction of the OS getting you to what you really care about:  the independent application workloads. Grantek can help you determine which is right for your operation.