Recognizing the Newest IBM Champions
Rob McNelly also shares IBM Support tips and details on recent eezh workarounds
“It is with great pleasure I can announce this year’s IBM Champions! We had an overwhelming number of submissions during our nomination process over the fall of 2023, and I’m happy to say that the IBM Champion program has experienced a 38% growth this year! This means even more IBM technology experts (who are non-IBMers) are going the extra distance to advocate for IBM...
“This year’s 1155 IBM Champions represent nearly 60 countries and are poised to make 2024 an outstanding year, as 63% return to the program, and 37% are newly recognized. Nearly 200 already have expertise in AI and watsonx.
“IBM Champions signify more than a prestigious name. These advocates are working with IBM technologies to innovate, to educate, and to advance their profession and business market. That 38% increase doesn’t even count our IBM Rising Champion Advocacy badge earners...We are excited to celebrate the launch of this new program, which can provide a pathway to becoming an IBM Champion. We issued our first batch of those late in 2023 and will continue to issue throughout 2024. Congratulations to the first round of IBM Rising Champion Advocacy badge earners."
Seeing this program grow so much is something I would have never imagined back in 2011. I'm so proud to be a part of it.
eezh WorkaroundsHere's some good news from Matthew Opoka regarding eezh. Thanks to a group effort, developers have come up with some creative solutions for dealing with issues that were introduced with recent HMC updates. I learned that all uses of sed were removed, with egrep and printf being deployed in multiple line substitutions. In some cases, it took 10 or more lines to replace a single sed statement.
Check the code for examples. Comments have been inserted to more easily track the changes. There's more in this discussion. This is another example of people sharing information and coming up with solutions that benefit everyone. It's truly impressive how quickly these workarounds were implemented.
Tips From IBM Support
1. VIOS bootlist limitations:“I installed VIOS on an IBM V7000 SAN disk (hdisk0). I added a second SAN disk (hdisk1) to rootvg and successfully mirrored the volume group. However, after mirrorios completed successfully (no errors), subsequent listing of the bootlist only includes 1 of 4 paths to the second disk. Why is that?
“Cause: pSeries systems firmware only allows a maximum of 5 devices (hdisk) or paths in the boot list. This is a limitation.
“Answer: Given the system firmware limitation, starting in VIOS 2.2.0, an enhancement was added with IZ72620 for administrators to be able to choose which path IDs to include in the bootlist.
“To specify 2 paths IDs per disk, run:
$ bootlist -mode normal hdisk0 pathid=0,1 hdisk1 pathid=0,1 $ bootlist -mode normal -ls hdisk0 blv=hd5 pathid=0 hdisk0 blv=hd5 pathid=1 hdisk1 blv=hd5 pathid=0 hdisk1 blv=hd5 pathid=1 $
2. Checking for non-AIX Toolbox packages:“Question: The Get Started with the AIX Toolbox for Open-Source Software page states:
“Non-AIX Toolbox packages that deliver files available in AIX file sets or AIX Toolbox packages might result in unexpected behavior. It is recommended to remove those non-AIX Toolbox packages.
“How can I determine whether my packages are from the Toolbox?
“Answer: You can check all your packages with:
# rpm -qai | egrep "Name|Install Date|Build Date|Build Host"“If you do not see IBM build hosts, you should consider reinstalling any packages that are available on the AIX Toolbox. This will help avoid package conflicts by managing package updates with normal DNF maintenance functions.
“Example of an IBM Build Host:
Build Host : aixoss2-lp3.pok.stglabs.ibm.com
3. Configuring TCP/IP routes with multiple interfaces:I realize that I usually stick to my lane of AIX and VIOS topics, but I found this information useful when working with IBM i clients that use VIOS for networking. One client that had a requirement to put multiple IP addresses in the same subnet onto the same adapter was experiencing difficulty getting everything to work as expected.
A ticket was opened with IBM Support, with the client receiving information on allowing TCP/IP packets to go in and out on the same interface:
“Problem: This document contains the steps to configure TCP/IP routes to allow TCP/IP packets to go out the same interface they came in.
“Resolving The Problem: Packets coming in a certain interface may go out a different interface depending on the routes configured.
“To have inbound traffic that comes in one interface and goes out the same interface, configure multiple redundant routes, each bound to a different interface.
“This will help inbound-initiated traffic only (when the IBM System i system is the server).”
View the link above to see the TCP/IP display route.
Some of you may support IBM i, or you may be a VIOS admin supporting IBM i clients on Power Systems hardware. While this is a scenario that few of you will ever encounter, I have an admittedly selfish reason to include it. Future me is bound to run into a similar situation; now I can return to this post and remind myself how to solve the problem. If a reader or two can benefit as well, so much the better.
About the author
Rob McNelly is a senior AIX solutions architect doing pre-sales and post-sales support for IBM Premier Business Partner Meridian IT Inc.
See more by Rob McNelly