Five rules help you properly control the scale of virtual machines

As long as we strictly abide by these five simple rules, our data center will be able to avoid the problem of over-expansion of virtual machines

Any technician who has tried to build a server virtualization environment must be impressed by the high flexibility brought by the virtualization technology. We often encounter tasks in our daily work, such as deploying a new application server, creating a complete multi-server development environment, or testing a new software package, with the help of virtualization technology, it usually takes only a few minutes to deploy the traditional physical environment. What's even more surprising is that creating a virtual environment is almost completely free: we do not need to declare a new hardware purchase order, nor do we have to wait for equipment transportation and installation, even the racks in the data room do not need to be disassembled or moved.

However, if everything has its own advantages and disadvantages. In many cases, the flexibility of Virtualization has become the culprit of unlimitedly expanding the number of virtual machines or even losing control. If this situation continues, virtual machines will inevitably bring about serious security and authorization problems-not to mention the unlimited demand for storage resources.

Fortunately, many merchants are aware of this kind of user demand. At present, a series of related management products have emerged in the market to solve the challenges in the field of Virtual Machine lifecycle management. However, if your company is small and there is no extra money (or time) to take care of these management tools, what should we do? Don't be afraid. I hope the following five tips can help you solve your troubles.

Rule 1: virtual machines are not free of charge

Although we can all deploy a virtual machine in the twinkling of an eye, it turns out that they are not free. The occupation of computing increment and storage resources will inevitably increase the operation cost, but this is only a small part of the additional expenditure. In addition, the increasing use licenses and other supporting tools of the operating system will also surprise managers who finally see bills-anti-virus software and security mechanisms, virtual machines are definitely a good tool for pig. To ensure the normal operation of the Virtual Machine System, we need not only to expand enough storage space for it, but also to equip it with an efficient backup system.

As long as the maintenance and operation items of the Virtual Machine System are listed in Excel, we can quickly create a reasonable and accurate cost estimation model. Through rigorous statistics and computation, we will find that virtual opportunities bring about a large amount of extra money. This result is confident that we will be able to persuade corporate decision makers who blindly respect virtualization technology. If the current resources are quite tight and users insist that technicians should use the virtual environment to solve their actual needs, this cost list will also become a powerful weapon for educating them-facts have proved that, fast and convenient.

Rule 2: do not always use "temporary" as an excuse

Combined with my personal work experience, the most serious area of Virtual Machine expansion is data testing and development environments. Thanks to virtualization technology, technicians can easily create a complete set of application environment copies, therefore, we often see that employees are very happy to use these environment copies to provide rehearsal venues for new software versions, updates, or training courses.

Whenever I remind everyone to buy or plan authorization licenses for various types of system backups, they always reject my suggestions on the grounds that "this is only a temporary solution, Don't worry. In fact, in some cases, such "don't worry" attitude may lead to severe consequences such as fines or legal sanctions.

However, more often, these systems gradually change from "temporary" to "long-term" over time ". Several years ago, a technician stirred up a copy of the environment based on the last two weeks of use, and today this set of things is still frequently used-it has no authorization or security protection, the original virtual environment has become a time bomb that enterprises cannot ignore. There are many reasons for such a situation, but we must grasp the bottom line and make correct judgments based on our own experience and knowledge. From this example, we should learn how to avoid deploying virtual machines in any "temporary" plan-virtualization technology should only serve the permanent plan, which is a principle Principle principle that most people understand but fail to do.

Rule 3: Specify the owner and owner

Whenever a new system is created, the most important thing is to find out who is responsible for the operation of the system. By assigning a unique owner for each system, which can be a system application analyst or a direct service object of the system, we implement virtual machines to specific owners one by one, in this way, when the production status of the system changes (including development, testing, production, recession, and final elimination), we can always put the system on a well-managed operating track. However, if there is a lack of actual owners, we may not know whether a virtual machine system needs to be backed up or whether it can be directly stopped after many years.

Rule 4: Create naming rules and stick to them

With the continuous expansion of the virtualization environment in enterprises, you should pay close attention to how to name and classify your own virtual machines. Assuming that you only have 10 or 20 virtual machines to manage, it is not difficult to track the status and distinguish them from each other. However, when the number of virtual machines reaches 40 or 50 in a few years, are you still clear about the situation? Even as a manager, we can make a lot of virtual machines well organized, but other employees will surely feel overwhelmed by our work (or new successors who will join us as the scale expands. To avoid this situation, we must formulate strict naming rules and stick to them.

If you are a VMware vSphere user, do not forget to use the field definition feature in vCenter, which makes it easier for us to track our own virtual machine information. In this way, I can successfully find the user touchpoints, creation date, estimated shutdown date, backup solution/progress, and even vendor support information that I need in a large number of virtual machine systems.

Rule 5: systems not used for more than thirty days must be cleared.

In the virtualization environments of most medium-sized enterprises, we often find many closed virtual machine systems. The reasons for this situation vary, but they are usually caused by application upgrades or elimination, and these idle virtual machines are often not cleaned up in time.

Among the above opinions, this situation is the most dangerous and has the worst impact. First, because the system has been idle for a long time, no one may remember its original role, who is in charge or whether it needs to be enabled-in a few cases, it is always forgotten in the corner. More importantly, because the system is already offline, it cannot accept OS patch updates or anti-virus software upgrades. With such a security risk in the machine, we are likely to introduce fatal attack activities on the network into the enterprise without knowing how to use old devices.

Therefore, the best way is to clean up the systems that have been idle for more than thirty days-no matter what their original role is or whether they will be used in the future. If you want to leave a backend for possible needs, consider using free tools such as VeeamZIP launched by Veeam. VeeamZIP can create a set of archive copies for the virtual machine system, but it also ensures that the copy cannot be enabled freely in the virtual environment and main storage reservoir. All in all, retaining a system that remains idle for a long time will not only cause waste of resources, but also pose a potential threat to enterprise business security.

Regardless of the size of the enterprises currently in effect and the degree to which they will develop in the future, it is essential to carefully handle VM creation and develop a suitable life cycle management cycle. These benign policies seem to have little effect in the short term, but over time, lazy management will eventually overload your storage network-a list of test servers that have been idle once, or a license audit list suddenly sent by Microsoft (this is not an alarmist, I have already met this situation many times.) It will immediately kill you to the ground. Never regret at that time Why I didn't develop a good habit of diligence and prudence.

Original article:

Http://www.infoworld.com/d/data-explosion/5-tips-vanquishing-virtual-sprawl-195695

Original article title: 5 tips for vanquishing virtual sprawl