IBM Reveals Changes to Power11 Partition Scoring
Occasionally readers will ask about documents that are linked in some of my older articles.
As an example, I was recently asked to locate a copy of the IBM document I’d linked to in this 2016 blog. If you’re at all tempted, don’t bother clicking that first link; the doc is no longer available.
The reality is that helpful information that one can find online may not always remain online. (This is one reason I launched my own archive.) It’s a frustration of the Digital Age, and it happens all of the time. It’s great that we no longer have to maintain volumes of books and paper printouts, but as easily shareable and accessible as links are, they’re also subject to change or deletion over time. This can be distressing and frustrating as the referenced information may remain useful long after its erasure from the internet.
With the tenuousness of online information in mind, let’s dig into some interesting and very much working links from IBM Support.
Security Notice Impacts Power10, SR-IOV Adapter Users
Via Chris Gibson, this recent security notification from IBM Support could impact users of Power10 systems:
“We have recently discovered an issue where Power10 systems with SR-IOV Shared Mode adapter configurations may encounter an incomplete state on firmware update to FW1060.50 or FW1050.50. A power off from ASMI and downgrade to previous firmware may be needed.”
IBM assesses this as a HIPER (High Impact/Pervasive) risk. The link covers multiple scenarios. If you are using SR-IOV adapters you should take the time to learn more about this.
Scale-Out LC System Log Collection Tools and Utilities
Also from IBM Support, a look at scale-out LC system event log collection tools and other utilities:
“If a Linux partition is up and running, then the standard sosreport should be collected in addition to running the plc.pl data collection script or the OpenBMCtool.
“The sosreportneeds to be run by root user, the script prompts for additional information and when complete will show the location of the output file.
“To collect the data from the BMC, the plc.pl script file (contained in the plc.zip package below) or the OpenBMC tool should be used by the customer to collect additional data.
“The plc.pl script is used for MTM: 8348-21C, 8335-GCA, 8335-GTA, 8335-GTB, 8001-12C, 8001-22C, 8005-12N, 8005-22N, 7063-CR1, 9006-12P, 9006-22P, 9006-22C, and 5104-22C.
“The OpenBMCtoolis used for MTM: 8335-GTG, 8335-GTC, 8335-GTW, 8335-GTH, 8335-GTX, 9183-22X, and 7063-CR2.”
Multiple tools and prerequisites listed, starting with plc tool installation:
“The Linux system used to perform the data collection must have the following tool packages installed:
ipmitool
perl
sshpass: for POWER8-based systems only (8001, 8005, 7063, 8348, 8335-GTA GTB GCA)
“The Linux system must have network connectivity to the BMC.
“The IBM OpenBMC tool is used to interact with the new OpenBMC on the 7063-CR2, 8335-GTG, 8335-GTC, 8335-GTW, 8335-GTH, 8335-GTX, and 9183-22X products.
“The scale-out LC system VPD update tool is used to make VPD modifications to Power Systems from the host OS of the machine.
“If a failure is seen when using the latest 3.3 version try using the 3.2 version, or update firmware to the latest and then try the 3.3 version.”
There is much more, so be sure to check it out.
Dealing With VIO Server Issues
This IBM Support doc covers what to do if your VIO server crashes/hangs and you need to determine the root cause:
“Symptom: VIOS hung (lost console and network access) or experienced a system crash.
“Cause: There are different factors that can lead a VIOS partition to crash or hang. This may include, but it is not limited to, code defect or hardware malfunction. Memory allocation to the VIOS LPAR plays a very important role in the VIOS availability. More often than not, VIOS crashes or becomes unresponsive (hangs) due to insufficient memory resources allocated to the VIOS.
“It is strongly recommended to read this document entirely before taking any actions to determine which scenario may be best applicable to your situation.
“Environment: VIOS version 3.1 and 4.1.
“Diagnosing The Problem: In order to determine what happened, system dump data must be collected immediately after the VIOS crashed or right after forcing a system dump when the VIOS partition is deemed to be hung.
“To determine if there is a good system dump, login as padmin and run:
$ oem_setup_env
# sysdumpdev -L
0453-039
Device name: /dev/lg_dumplv
Major device number: 10
Minor device number: 11
Size: 21010432 bytes
Date/Time: Tue Feb 12 13:05:33 CST 2015
Dump status: 0 <—–
dump completed successfully <—–
Common dump status codes include (but are not limited to):
-0 Successful system dump
-2 Dump device is too small (partial dump may be available)
-3 Dump facility has failed (dump crashed or did not start)
-4 Dump failed due to i/o error (bad disk or disk adapter)
The doc goes on to explain how to resolve this issue.
Changes to Power11 Partition Scoring
Here’s an interesting look at Power11 minimum affinity scoring from IBM’s Devesh Singh:
“With the introduction of Power11, two new attributes have been added related to partition scoring: a minimum desired affinity score and an action to the taken at boot or for Live Partition Mobility (LPM) when the current affinity score is less than the minimum specified score. The actions supported are to ignore the low score, warn at boot/LPM or fail boot/LPM.
“Affinity score: PowerVM uses a 0–100 affinity score to rate how well a partition’s processors and memory are aligned with respect to NUMA (Non-Uniform Memory Access).A score of 100 means the resources (CPU and memory) are optimally assigned.If resources are spread across more chips, dual chip modules (DCMs),or CEC drawers than required,the score is reduced based on how far they deviate from the ideal placement. Lower scores indicate less efficient resource locality, which can affect performance.”
IBM TechXChange Reminder
TechXchange is coming October 6-9 in Orlando, Florida:
“The countdown is on for IBM TechXchange 2025—the premier event for technologists, developers, architects, and innovators shaping the future. Whether you’re driving digital transformation, modernizing your infrastructure, or harnessing the power of AI, TechXchange 2025 is where ideas turn into action.
“One track you won’t want to miss: the IBM Power Track—your gateway to mastering AI, automation, hybrid cloud, and modernization on one of the most trusted enterprise platforms in the world.”
The Challenge of Something New
There is still time to sign up for the Flagstaff to Grand Canyon Stagecoach Line races and relays. The two-day event (September 20-21) features a 100-mile run, 55k run, and 100-mile mountain bike race and relay.
While I plan to be at the finish line to cheer on the racers, I will not be participating this year. Instead, I am focused on a sprint triathlon that I will compete in on September 13.
I do fully understand that most of you have neither the time nor the inclination to attend some niche race in northern Arizona. And of course, that’s fine. But I mention these things for another reason. My hope is that it may spark an idea in some of you to find and take on some fun, unique activity in your own area. Maybe it has nothing to do with hiking, biking or running, but it’s something you’d like to try and it poses a challenge all the same.
On that note, feel free to reach out and share the types of challenges you prefer to tackle. Maybe you’ll inspire me to try something new. Why not? After all, I’m not getting any younger.