Freelance virtualised: How (available) would you like it?

Whether replacing dozens of process stations with a redundant server environment with central management or clear monitoring while simultaneously increasing availability: there are many reasons why virtualising an ABB Freelance system is worthwhile. We show you what options are available to you and why a Freelance upgrade to version 2019 is recommended in the course of virtualisation.

Graphic to visualize the relations between controllers, 2 cluster hosts and a central storage. — The most minimal solution for an HA cluster consists of two hosts and a central storage.

About ten years after a wave of virtualization swept through and revolutionized office IT, it is now rightly spreading to automation systems The advantages are clear. A large number of systems, such as engineering and operator stations, can be run centrally in a server structure as —hardware-independently and cost-effectively. Advantages:
• Centralized management
• Reduced maintenance effort
• More efficient hardware utilization
• Hardware replacement without interruption
• Centralized backup solution
• Engineering and test environments that can be provisioned very quickly

Freelance + Virtualisation = Freedom

The “Freelance” is still recommended. Relatively simple and very robust, it fulfills its purpose as a central control level for distributed systems. A minor drawback is the lack of redundancy at however, this can be virtually eliminated through virtualization. Which is better?

Over the past 30 years, CTE has designed, engineered, and implemented numerous and control systems. Products from Siemens, ABB, Emerson, GE Fanuc, Rockwell, and Copa-Data were used in these projects. More important than the question of whether one is better than the other is what the requirements are and how the system can be planned and implemented. Even the best control system can end up in a confusing mess due to improper planning. On the other hand, a structured SCADA project can offer a customer and their facility exactly what is needed.

Variants of the cluster structure

To ensure the system runs without interruption, a High Availability Cluster (HA cluster) is required Whether using “VMware vSphere” or “Microsoft Hyper-V,” both vendors support the key features of a cluster: HA, live migration, and automatic load balancing. In an HA cluster, virtualized systems are not run simultaneously on both hosts; instead, if one fails, they are restarted on another available host. Restarting the machine causes a brief outage of a few seconds. It is also possible to run a VM on two hosts simultaneously as a Fault Tolerant (FT) However this solution involves high hardware and licensing costs on the one hand, and on the other hand, it does not protect against a system failure. Therefore, it is rarely recommended from a cost-benefit perspective.

Quite traditional

The minimum solution for an HA cluster consists of two hosts and a central storage (see Fig. 1). In the best case scenario, the storage hardware is designed to be completely redundant so that if a component fails, the VMs on the hosts can continue to run. With this variant, one cluster host is allowed to fail and the VMs are restarted on the second host. However, if the shared storage fails completely, VMs can no longer run on the hosts.

Even smarter

Another variant is a modern hyperconverged HA cluster, which no longer uses traditional storage, but combines the local disk in the individual hosts into a logical volume with the desired reliability (see Fig. 2). A major advantage of this solution is that the entire central storage can be dispensed with. Thanks to the local disks in the hosts, there is no longer a "single point of failure". One component can fail completely, regardless of which one. However, this variant requires at least three hosts and a 10G network.

Even more independent

The premium version with full site redundancy consists of a hyperconverged HA cluster with at least four hosts. The failure of a single host or even an entire site is automatically intercepted by the cluster and the VMs are available again within a short period of time. Central and cross-location services can thus be provided regardless of location. At the same time, the VMs can be brought close to their assigned field components in order to create cells that are more independent of external circumstances.

However, PLEASE NOTE!

Clustering 'only' protects against the failure of individual hardware components, but not against software problems. Targeted monitoring of the relevant systems is therefore important for such infrastructures, in order to provide immediate alerts in the event of redundancy failures and to proactively counteract problems that can be identified in good time. Clustering also fails to protect against changes like reconfiguration, deletion, or encryption within a VM. However, the cluster consistently manages these actions across all instances. A professional backup solution is therefore always worth its weight in gold.

CONCLUSION - When is something worthwhile?

Despite all the advantages, virtualisation is often a question of cost. If many operator stations can be replaced, the acquisition costs of the server structure are already lower than the costs of the individual hardware. However, if the reduced maintenance costs and higher availability are also taken into account, the decision is often made in favour of virtualisation even if the hardware costs are somewhat higher. When replacing a small number of clients, we believe it should be carefully considered whether virtualisation is worthwhile. After all, the greater complexity of the system landscape should not be underestimated. Unless the servers already exist or are to be used for additional tasks. It's up to you to decide which option to choose.

Combination with Freelance Upgrade

In addition to new hardware and optimised user operation, Freelance 2019 offers two new features in particular, which can be very helpful. Firstly, up to four monitors per operating station are now supported and secondly - particularly useful in connection with virtualisation - the user administration can be connected to the Windows domain controller.