A Warning About Running ezh and eezh Scripts in Updated HMC Environments

In 2020 I wrote about a couple of scripts called ezh and eezh:

“EZH is a script for the [HMC] to provide an alternate, easier to use, command line interface for many common commands, and the goal of the project is to make the HMC command line interface easier to use for day to day administration tasks.

“…Enhanced EZH script is a fork of EZH by Brian Smith for the [HMC] to provide an alternate, simpler command structure which makes the HMC command line interface easier to use for day to day administration tasks. The vision for this project is to enhance and bring EZH commands forward and inline with technologies provided in the newer HMC code base.”

For those of you who use these scripts, you should know that they’ll break the next time you update your HMC. Matthew Opoka, who maintains eezh, gave me this heads up. He explains that, in addition to ezh/eezh, the latest patch will break any HMC script that utilizes sed, since support for the sed command is basically being removed. (These PTFs are intended to address CVE-2023-38280. See the PTF MF71298 and MF71299 readme files for details.)

It’s possible this feature could be reinstated in the HMC code through the IBM Request for Enhancement (RFE) process. If you use sed with the HMC, I suggest you register on IBM’s RFE portal and share your opinions. For the uninitiated, the RFE community is well worth your time and participation, as it gives all of us the opportunity to help shape the future of Power Systems development.

VIOS to NIM Mapping, Recent AIX Enhancements, IBM PowerDraw

These items come courtesy of IBMer Chris Gibson‘s mailing list.

1. This recently updated IBM Support doc covers VIOS to NIM mapping. Check the link to see the master levels needed if you’re using NIM to backup, install, or update a VIOS partition.

2. Here’s an extensive list of recent AIX enhancements, including AIX 7.3 Standard Edition TL 2 and AIX 7 Enterprise Edition 1.10.

3. And here’s a description of a new configuration tool called PowerDraw, which provides an interactive, graphical representation of Power Systems environments:

The product uses the included capture module to collect all the information about the server, PowerVM (VIOS), and partitions from the HMC into a single text capture file… Any modern browser can display and interact with these diagrams, allowing [clients] to easily explore the relationships with mouse over highlighting of connections, and popup boxes that display all the detailed information. It can be useful as the starting point for problem determination, or for easy evaluation of virtualization best practices. The SVG files can be integrated into browser-based dashboards or used as stand-alone local files viewed by a browser. No web server is required.”

Incidentally, Chris’s mailing list is open to all. Email him if you’d like to be subscribed.

More AIX Problem-Solving Tips

Here are some other things I’ve recently come across.

* This IBM Support doc describes how to use MTU 9000 (jumbo frame) in an environment with AIX LPAR and VIOS.

* Here’s an oldie but a goodie on preparing a Power server for a concurrent SAS or SSD disk replacement in AIX or VIOS:

“It is important to first determine disk array configuration including protection level and then follow appropriate procedure. To perform these procedures AIX root access is required. On VIOS run “oem_setup_env” to switch from user padmin to root.

“If you already established your level of protection, you can go directly to the Procedure.

• Procedure 1: Replacing an AIX System Disk, hdisk JBOD (Just a Bunch Of Disks)
• Procedure 2: Replacing a pdisk that is part of a RAID0 unprotected array with a single pdisk
• Procedure 3: Replacing a pdisk in a RAID5 or RAID10 protected array
• Procedure 4: Replacing a pdisk in a protected RAID array that was previously a Hot Spare
• Procedure 5: Replacing a pdisk in an unprotected RAID0 array with multiple pdisks
• Procedure 6: Replacing a pdisk in a RAID6 array
• Procedure 7: Replacing a hot spare disk that is either failed or has a PFA (Predictive Failure Analysis) indication”

* Here’s another useful nugget. Use this HMC command to determine the MAC address of the virtual ethernet adapters on each LPAR:

lshwres -r virtualio --rsubtype eth --level lpar -m <managed system>

* And lastly, a troubleshooter from IBM Support. An AIX system rebooted, but getty didn’t start, preventing a console login:

“The most common reason for getty not starting is failure of a preceding line in /etc/inittab. Since that line does not complete, the line which starts getty is not read….

To troubleshoot:

Check the console setting

# lscons
vty0

Check the state of vty0

# lsdev -C|grep vty0
vty0 is available

Check process table for getty

# ps -ef|grep getty

There is no output, which means that the getty process isn’t running.

Check to see what process is prohibiting getty from starting or respawning.

# who -p

“Usually, the culprit is the last process in the who -p output. In this example, it is l2, which is called from the /etc/inittab file. Note: l2 is causing the system to not completely read /etc/inittab.”

See how they addressed the issue.

---