IBM: Be careful what you say

IBM is once again demonstrating that it can be its own worst enemy. This time it involves their new mainframe system, the z13. IBM has a lot of things to be proud about with the z13, but their marketing messaging is definitely not one of them.

IBM is rightfully very interested in promoting their mainframes as being a good choice for running private clouds. They've spent a lot of money contributing to things like OpenStack as well as internal software development efforts to make the mainframe a real contender. IBM has also been investing a significant amount of effort in KVM for the mainframe. The thought there is that new-to-the mainframe customers might not want to install z/VM immediately for their virtualization needs, but they might already be familiar with KVM on Intel/AMD. With a lower barrier to entry in terms of skills, they hope to find it easier to finalize sales with these customers.

Unfortunately IBM has conflated the two topics in their marketing materials, resulting in a very misleading picture of the z13's capabilities when running KVM. There have been a good number of media and blog references to IBM's claim that up to "8,000 virtual machines" can be run on the z13. Usually, the same slide or advertisement also talks about KVM.  (At the time of the z13 announcement, IBM also released a Statement of Direction that they intend to provide their own KVM offering for the z13.)

What IBM has not made clear at all is that the claim of 8,000 virtual machines is based on running z/VM, and not KVM. People outside of the z/VM community are not aware of this and are assuming that they can get those same results with KVM.  With today's state of the art in KVM, that's just not possible on a z13.

Various market analysts/reporters repeated this mistake at the time of the announcement (see a couple of examples below). And now IBM's business partners are starting to propagate the same incorrect information.

For whatever reason, IBM seems extremely reluctant to even mention the z/VM product by name in public. This, despite it being a significant factor in the great success Linux has had on the mainframe.  Apparently IBM has been called on this multiple times, the most recent at the SHARE conference in Seattle, but haven't changed their marketing materials. This is really sad, since z/VM is overall probably the best hypervisor in the industry in many different ways. It's something IBM should be shouting about instead of seemingly trying to hide from the market.



http://virtualizationreview.com/articles/2015/01/16/ibm-says-mainframe-not-dead.aspx
http://www.drunkendata.com/?p=4688

Advantages of Working From Home

One of the advantages of working from home is that you come to appreciate just how many times a day your dogs prevent attacks from outside the house by evil air molecules. Apparently they're afraid of barking dogs and run away.

Things I learned about leadership from my pit bull

As Cesar Millan has been known to say "You don't get the dog you want, you get the dog you need." When my wife and I adopted two rescue dogs from local shelters, that turned out to be true for me. While I've been in leadership roles before at work as a team leader, I didn't always feel comfortable with parts of the job. Trying to be a responsible owner of a pit bull has changed that to a large extent. While I already knew some of the things I've listed below, they were definitely reinforced for me by working with Copper and Kashmir. I'm sure this list isn't complete. My dogs have lot more to teach me.

• If you're supposed to be the leader, you need to act like it.
• If you don't lead, someone else likely will. You probably won't like what they do.
• Leadership first; friendship/affection/love later.
• Don't be apologetic when giving orders or direction. It's part of your job as a leader.
• Don't be harsh when giving orders.
• Be very clear on what your expectations are, and what rules you have.
• Be consistent in enforcing the rules. Sending mixed messages will cause confusion and demoralization.
• Don't have more rules than truly necessary.
• Trust in you and respect of you by others has to be earned. Trust and respect of others by you should be the default until shown otherwise.
• Rewards and recognition have to be individualized or you risk them being meaningless to the recipient.
• You have to know the recipient well enough to know what reward is meaningful to them. Take the time to find out.
• Rewards for achievement need to be given immediately, not some time later. Plan ahead.
• Never give praise or rewards when they haven't been earned.
• Always give praise or rewards when they have been earned.
• Praise should be effusive. Discipline should be calm, restrained, and no more than necessary.
• If a reward was promised for an achievement, you have to deliver it, even if it costs you personally.
• Don't promise rewards if you're not sure you can deliver them.
• Making a big deal out of failure is counterproductive. Deal with it, learn from it and move on.
• Preventing bad behavior is far better than handing out consequences afterward.
• When bad behavior is exhibited, it needs to be addressed immediately. Ignoring it will only make things worse.
• When things go haywire, keeping calm is by far more effective than getting excited or panicked.
• If they truly believe you are the leader, and that you "have their back," they'll do nearly anything in the world for you.
• When you're teaching a new skill, patience and calm persistence is critical. Achievement will likely come in small increments, not "all at once."
• If you don't know how to teach a skill, don't try. Find someone who does know how to teach it.
• Skills need to be periodically practiced and reinforced, or they'll be lost.
• Failing to provide training is a recipe for disaster.
• Fairness is often in the eye of the beholder. You may need to go a little out of your way at times to help everyone feel that what you're doing is fair.
• Your needs are not the only ones that matter. If their needs aren't being met too, trouble is inevitable.