Reader Feedback On IBM i Wish List For 2015
January 26, 2015 Hey, TPM
First, my disclaimer. I have worked my whole life in the S/3X space: 31 years as an IBMer, last 10 working for the (now) IBM i brand and Rochester, then nine years working for IBM Business Partners. Next, my wish list (biased by the above): 1) I’d like to see graduated pricing for IBM i, regardless of software tier. One of the biggest impediments to sales of new footprints is Initial Purchase Price. P05 licenses of IBM i and its LPPs are very competitive. P10 is expensive and P20 and above are prohibitive. I’d like to see IBM make the first license of IBM i cost $2,500, regardless of tier. The second $5,000, the third $7,500, and so on in some stair-step fashion. It is fair to assume that machines with more processors are doing more work, and that the OS should cost more as the machine does more. SWMA could remain based on tier. But the price of the OS should be more like AIX. Granted, IBM i includes DB2 and WebSphere Express and some other software that AIX does not. Maybe IBM could add a DB2 surcharge for each license. But with P10 machines that have north of 140K CPW, there is no sense in pricing the machines out of the market with OS licenses that cost $44K and more. 2) An annual amnesty period for After License Fees (ALF) on expired SWMA. We have lots of older machines still lying around. A major impediment to getting them current is the ALF, which runs up to 30 percent of the price of the SWMA. You can say that they should just buy net new licenses, but at $15K per OS, it is very expensive for many of them. Conventional wisdom is that you don’t want customers to be able to just drop SWMA, and renew when they want to upgrade. But you also lose access to PTFs and Support Line. You lose the opportunity to transfer OS licenses for a $5K fee. In the end, you want to make it easy for customers to get back under SWMA, not hard. 3) Remove the marketing restriction of no loops on a P05 machine. The only reason for wanting more loops anymore is to get additional DASD. And you can accomplish that with Fibre Channel adapters and a small V3700 SAN. But the SAN is more expensive to acquire, install, and maintain. Limit the number of drawers on the loop. Power8 allows a lot of drives in the CEC, but makes it hard to partition the machine if you want to have physical LPARs rather than virtuals. 4) Make the media charge for IBM i reasonable. I understand that free was too cheap. But $350 for a set of DVDs is stupid. And most customers don’t have the time or skills to go to an IBM website and download the media to optical image catalogs. 5) Simplify the process of getting keys for CBU machines. It makes no sense for the business partner to have to go look up every LPP and feature that the customer has, then have an IBMer send in the stupid Universal Backup Form. The partner should be able to simply send in the primary and CBU serial numbers to the Key Center, and they should generate the two year temp keys for all of the entitled products on the primary. 6) Bring back serial number upgrades. Power8 does not allow them. Power7+ only allows a few. Make it easy for a customer to migrate their software to newer hardware. You can argue that the $5K transfer fee makes it easy, but $5K is still $5K that could be spent on something else. Let customers upgrade anything from Power5 up to new machines. 7) Find a way to bundle the HMC with a new system to reduce the price for a Linux workstation that people use of nothing after the initial install, other than maybe PTFs. New HMCs cost $10K. That is dumb. LAN console is OK, but not as good as HMC, and HMCs are required for LPAR’ed machines. Make them reasonably priced. That’s my quick list. Good luck working on them with IBM. –Doug Now that, Doug, is what I call a very precise wish list. Thanks for telling us your thoughts. Anyone else want to get it off their chest, hit that contact button above and let us know. We need all of the good ideas we can find. –TPM RELATED STORIES IBM i Predictions For 2015 From Around The Community
|