I helped a friend move his belongings into a new house last weekend. We had a group of friends, and I remember the difficulty we had in choosing the best weekend to get together. Everyone had an excuse not to come, from responsibilities to children, to a family celebration, to vacation and various other reasons. We were far from flexible and elastic.
This type of inflexibility is a common problem in IT infrastructure as well.
The number of virtual machines is increasing in each physical server, so it becomes very difficult to completely stop a machine. Indeed, each virtual environment is often a service or a very specific project in a company, and it’s a big challenge to get numerous departments or IT services to stop, just as it was difficult to get many people together at the same time to help my friend move.
Particularly in Unix hosting in critical environments, you must obtain the agreement of all projects before stopping the machine. Choosing the hour to stop it often requires lengthy negotiations, and in severe cases this can lead to a deadlock.
In the IBM Smarter Computing approach there is a simple device to facilitate the work of stopping a physical server, though people often don’t know about this feature and how it works. It is called Live Partition Mobility (LPM). LPM best practices and rules are described here: IBM PowerVM Live Partition Mobility: An IBM Redbooks publication.
The interest of LPM is to move “on the fly” any activity running AIX in a partition. This concept is particularly useful when using partitions that include critical databases. The user will not detect a change of server (just a “freeze” of one or two seconds). Here is a video example:
This device is used in two major cases:
- Moving servers between different rooms or different sites (like my friend who moved into his new house)
- In this case, just plan ahead, leaving enough space on the new machine to move the production load.
- A rise of critical firmware
- In this case it may be advantageous to shift the burden of production onto another machine and test environment before putting it on the production server up to date.
- This way backtracking is possible.
We are at the beginning of this flexibility of moving loads “on the fly.” This capability is a key approach in cloud computing.
LPM already lifts many limitations (for example moving from blades environments to other types of machines). The new machines like IBM PureSystems incorporate these features. No doubt, important progress will be made in the near future in this arena.
Philippe Lamarche is currently an IBM Systems Architect in the hardware division (STG) since 1995, working with french industry customers and System Integrators. He has spent over 30 years at IBM in different technical positions. As a presales technical role he is a Certified IT Specialist at expert level. You can reach him on Twitter: @philip7787.
To effectively compete in today’s changing world, it is essential that companies leverage innovative technology to differentiate from competitors. Learn how you can do that and more in the Smarter Computing Analyst Paper from Hurwitz and Associates.