Hertvik Business Services – Joe's Blog

This is the stuff that makes me crazy.

From Tim Prickett Morgan at IT Jungle.

To put it bluntly, IBM’s pricing on the new PowerLinux 7R2 rack-based server, which is only certified to run Red Hat Enterprise Linux or SUSE Linux, is crazy stupid lower than what IBM i and AIX shops have to pay. Or more precisely, to get the PowerLinux boxes within spitting distance of a comparable two-socket X86-based server (by Big Blue’s reckoning, not mine), IBM has had to radically cut the prices it charges for processors, memory, and disk capacity. The price reductions, which I figured out after drilling down into IBM’s pricing information, are jarring. It is a pity that the BIOS inside the PowerLinux 7R2, which is based on the Power 730 hardware, will not let either IBM i or AIX boot on them. And IBM better make sure that disk and memory features for the PowerLinux machines don’t work in regular Power 730 machines, too, or it will have just created a black market in parts based on these rather large price differences.

Sigh…

There’s a lot more revelations in Tim’s article. And you can read it all by clicking here.

**************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

From Jerome Hughes, if you want to find all kinds of images for IBM i for your next presenation, including the round blue “IBM i for business” logo and a whole treaure trove of other IBM i icons, do the following:

1. Go to Google and click on Images in the menu bar
2. Type in “IBM i” for business in the Google Search box
3. You’ll see an incredible array of images centered around IBM i

So go here if you’re looking for IBM i images.

IT Jungle and other publications are still sifting through some of the IBM i and and Power systems announcements made on April 24th.

Tim Prickett Morgan discovered that starting July 10th, IBM will start adding shipping and handling charges to selected Power i equipment. This surprises me because I didn’t realize that IBM wasn’t charging shipping and handling before this.

Tim has the details including pricing (what’ll you see what IBM is charging for shipping a Power 795 system). Click here to read the whole article.

**************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

On Tuesday, May 15, 2012, I gave a presentation to the Omni User group on Future Directions for the IBM i. I had a blast and the presentation went over very well. Special thanks to Lorraine Morgan, Bill Parks, Jerome Hughes, Tom Duncan, and the rest of the Omni User crew for having me speak. I enjoyed it a lot. Let’s do it again soon.

Omni and I are both posting the slides from my presentation on our Web sites. To view my slides, click here. If you have a user group or are interested in having me speak, drop me a line. I always like connecting with the IBM i community at large.

And if you’re in Chicago and haven’t been to Omni, you’re really missing something in the IBM i world. Check them out by clicking here. They have a great golf outing coming up (I’m thinking of playing) on July 19th and their annual Omni technical conference is on September 18, 2012.

**************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

IT Jungle just posted my latest tip on stopping your system from running the system startup program after a full system backup. Here’s the reader question I’m answering today.

Good article on the secrets of the IBM i IPL parameters. Is there a way to stop the system from restarting after my full system backups complete (GO SAVE, option 21)? There are times that it would be great for the system to stay in restricted state after a backup, but the startup CL program always runs. How do I stop it from running?

Click here to read.

**************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

I’ll be speaking at the Omni User Group in Oak Brook, IL tomorrow night (May 15, 2012) on the topic of What’s New with IBM i: Future Directions.

If you’re interested, I’ll post directions for downloading the presentation after the speech.

**************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

From various COMMON sessions that discuss Power VM and LPM and from some talks with IBM on the COMMON Expo floor, I pieced together the following info. IBM please feel free to suggest changes if I have anything wrong.

Here’s what I learned about Live Partition Mobility (LPM), the ability to move a running IBM i partition from one IBM Power machine to another. In addition to this information, Dawn May, Senior Technical Staff Member at IBM Rochester, wrote a very detailed piece about IBM requirements for using Live Partition Mobility, which can be accessed by clicking here.

• Live Partition Mobility will be controlled through the HMC.  Good thing they decided to keep it
• The two partitions (the source and the target) must have equal capabilities in terms of hardware or the transfer may fail
• If running 5250 sessions, the two partitions must have equal capabilities in terms of interactive capability, or the transfer may fail if there are interactive sessions running
• The larger the pipe, the quicker the transfer between the boxes
• There is a little bit of a grey area when the partition will exist on both boxes but it is fairly short
• Both boxes must have virtualized storage
• LPM occurs when the memory and CPU are transferred between the boxes and the receiving box is pointed towards the storage of the sending box
• LPM transfers can occur for a) running partitions; b) suspended partitions; and c) inactive partitions. It is not limited to active partitions
• IBM has not announced how it will handle transferring license keys between the two boxes. It’s unclear what state the licensed program licenses will be in when they reach the new box. There may be a good chance that when the licenses move to the new box, they will treated the same way as when you transfer licenses from a production machine to a Capacity BackUp system (CBU). ie, the licenses could revert to 70-day temporary licenses.
• It’s also unclear how LPM will handle license keys from third-party vendors. Third-party licenses are frequently keyed to the serial number of the box, the partition ID, and some other machine-specific parameters. You could be able to transfer a live partition only to find you still have to run certain procedures before the third-party software will work on the new partition, because the software won’t run on a new box with a different serial number, LPAR ID, etc.
• IBM isn’t planning on its users using LPM for HA or DR. It has uses for maintenance, load balancing, upgrading firmware, upgrading hardware, and other tasks but it’s not a great fit for HA/DR. For one thing, it’s dependent on shared disk between the two machines. If a disaster wipes out the disk from the source machine, there will be no recovery.
• You probably won’t be able to transfer partitions from a lower level P-Group box (say a P05) to a higher-group box (P10 or above). This is due to licensing. You can’t run cheaper licensed programs on a more expensive box.

That’s it for now. Please feel free to chime in with any other info IBM or non-IBM people may have found on LPM.

***************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

In an IBM COMMON session on Power VM, IBM gave a live demonstration of moving a running IBM i partition from one box to another using the new Live Partition Mobility and the HMC. Here’s how it went down.

• The IBM presentator brought up a green screen with the Work with System Status screen (WRKSYSSTS) on it
• He brought up a DOS command window that continually pinged the IP address of the partition with the WRKSYSSTS screen on it
• He ran through the procedure on the new HMC to move the box to another production box
• The WRKSYSSTS screen remained active and the command window successfully continued to ping except for 2-3 dropped pings when the transfer actually occurred

This was pretty impressive. I know there’s going to be some glitches but this looks like the beginning of a nice new feature that lots of us will probably be using in the next five years. We talked about the uses for LPM and came up with the following list:

• When you need to perform some kind of hardware maintenance on a partition and want to keep the machine running
• When you need to perform an upgrade without disrupting processing
• When you are working on the keck or adding another disk drawer or other process where you would normally take down the machine
• When you need to move the partition within a cloud to straighten out load balancing or point it towards another set of resources, such as other telecom lines on the second machine

IBM stressed that LPM wasn’t for HA or disaster recovery and that LPM was best used when the machines were on the same subnet (though they can be moved between subnets). But there may be a possibility that partitions can be swapped at a longer distance sometime in the future.

I see LPM the same way I’ve seen other technologies when they are first introduced. It may be a little kludgy at fist and it may take everyone a while to start using it but once we get started, it will probably be everywhere. But LPM has been available on AIX for a while now and it’s now ready for the IBM i.

I can’t wait for my first real need to set one up.

***************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

At COMMON today, I told someone the gory story of how in the past six months, I’ve had three occasions where two of the six IBM i partitions I manage have needed emergency IPLs to restart their systems.

My audience’s response was a simple but shocked “NO!!!!”

I responded that this was what happened and he replied again, “NO!!!!” as if he couldn’t conceive of such a thing happening. I told him a third time, and he again said “NO!!!!!”

I know how he feels. I can’t believe this happened either.

But it did. I actually experienced three IBM i emergency IPLs over six months, two of which occurred this year. And none of them were due to obvious hardware failures.

This is a level approaching the track records of a Windows server rather than the hallowed IBM i boxes where the machine’s reliability is not only legendary, it’s mythic. This just doesn’t happen in an IBM i shop, and I’m still scratching my head over it.

In this week’s IT Jungle, I put out the details of my three recent IBM i #FAIL situations and what I could have and maybe should have done differently. You can read all the details by clicking here.

***************************************************
Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.

In between looksoftware’s annual user conference and the start of this year’s COMMON at Disney Land in Anaheim, California, was an unusual event: the iBelieve revival meeting for IBM i evangelisation. Dan Burger from IT Jungle was there and reported on the passion, glory, and star-power of those speaking at the event and those who attended the event.

Per Dan:

The point of the event was not to convert any of the attendees, but to inspire them to help with the conversion of the non-believers. It contained equal parts pride and passion. And was designed to pick apart misconceptions about IBM i.

And iBelieve had IBM i-star power attached to it with such community luminaries as father-of-the-i Frank Soltis; IBM i Product Manager Alison Butterill; IBM i programming gurus Jon Paris and Susan Gantner; and of course, Trevor Perry. It sounded like a pretty intense party that I would have like to seen.

To get the scoop on what happened, check out Dan’s report on iBelieve by clicking here.

***************************************************

Follow Joe Hertvik on Twitter @JoeHertvik. You can also add Joe to your professional network on LinkedIn by clicking here.