Saturday 6 November 2021

Know Your Software Licenses - The right for every user

 With many cloud environments in operation today, you pay for resources by the CPU-hour. The cheapest virtual machine in the Amazon Cloud, for example, costs $0.10 for each hour you leave the instance in operation. If it is up for 10 hours and then shut down, you pay just $1.00—even if that is the only use you make of the Amazon cloud for that month.

In a real world, you might have the following operating scenario:

  • From midnight to 9 a.m., run your application on two application servers for redundancy’s sake.
  • From 9 a.m. to 5 p.m., launch six additional application servers to support business-hour demand.
  • For the evening hours through midnight, reduce the system down to four application servers.

Adding all that up, you pay for 110 hours of computing time. If you were using physical servers, you would have to purchase and run eight servers the entire time.

Unfortunately, not all software vendors offer licensing terms that match how you pay for the cloud. Traditional software licenses are often based on the number of CPUs. An organization that uses 10 application servers must pay for 10 application server licenses—even if 5 of them are shut down during the late night hours.

So, when moving to the cloud, you must understand your licensing terms and, in particular:

  1.     Does the software license support usage-based costs (CPU-hour, user, etc.)?
  2.     Does the software allow operation in virtualized environments?

Because the cloud makes it so easy to launch new instances, you can readily find yourself in a situation in which a lower-level staff member has launched instances of software for which you don’t have the proper licensing and, as a result, has violated your license agreements.

The ideal cloud-licensing model is open source. In fact, the flexibility of open source licensing has made the Amazon cloud possible.

 If you can remove licensing concerns altogether from your cloud deployments, you are free to focus on the other challenges of moving to the cloud. Although some open source solutions (such as Apache and most Linux distributions) let you do whatever you want, you may have to deal with licenses if you get supported versions of open source software, such as Red Hat Enterprise Linux or MySQL Enterprise. Luckily, these licensed offerings tend to be cloud-friendly.

Beyond pure open source, the best licensing model for the cloud is one that charges by the CPU-hour. As the cloud catches on, more and more software vendors are offering terms that support hourly license charges. Microsoft, Valtira, Red Hat, Vertica, Sun, and many other companies have adopted per-CPU-hour terms to support cloud computing. Oracle promotes their availability in the cloud but unfortunately retains traditional licensing terms.

Software that offers per-user licensing can work adequately in the cloud as well. The challenge with such software is often how it audits your licensing. You may run the risk of violating your terms, the license may be tied to a specific MAC or IP address, or the software license management system may not be smart enough to support a cloud environment and unreasonably prevent you from scaling it in the cloud.

The worst-case scenario in the cloud is software that offers per-CPU licensing terms. As with some per-user systems, such software may come with license management tools that make life difficult. For example, you may have to create a custom install for each instance of the software. Doing that impairs the flexibility that the cloud offers.

Some CPU-based software licenses require validation against a licensing server. Any software with this requirement may ultimately be inoperable in the cloud if it is not smart enough to recognize replacement virtual servers on the fly. Even if it can recognize replacements, you’ll have to make sure that the license server allows you to start the number of servers you need.

Unless a license server is going to be an impediment, however, the result is no worse than a physical infrastructure. If all of your software fits into this model, the benefits of the cloud may be small to nonexistent.


Post a Comment

Note: only a member of this blog may post a comment.

Machine Learning



Java Tutorial




C Programming


Python Tutorial


Data Structures


computer Organization