IBM Support

Quick Intro to the AIX Best Bits

How To


Summary

"AIX Best Bits" has two Support Articles:
1) The public vote results and
2) Quick Introduction (this article)

Objective

Nigels Banner

Find the AIX Best Bits votes:  Link to other support Article

Steps

Visits to the original website 11177

Quick Intro to the AIX Best Bits

image

PowerVM: 1st place

Briefly: Enterprise-class virtualization on the complete POWER range to run AIX, IBM i, and Linux.

Includes:

  • Hardware Management Console (HMC),
  • Firmware-based POWER Hypervisor called "pHype
  • Virtual I/O Server (VIOS)
  • Providing virtual machines (VM), virtual network & virtual storage

Benefits:

  • All you need to create & control of virtual machines including exact CPU cycles guaranteed and shared, and Memory and direct adapter I/O or virtual I/O
  • Dual VIOS provides redundant I/O paths and live VIOS updates with no loss of service.
  • With plenty of controls and options
  • High security - the hypervisor has zero reported vulnerabilities

Example use:

  • Run a virtual machine with precise control of the CPU, memory, and I/O
  • And control the sharing of resources for max performance and reduced software licenses

More information:


image

System Management Tool (smitty): 2nd place

Briefly:

  • Terminal-based menu & panels system for managing all of AIX and its subsystems

Benefits:

  • Quick, simple to understand & man-power efficient control of AIX
  • Reduce system admin errors
  • Command options are explained as you complete the form
  • Teaches the underlying command used for later shell scripting
  • Comprehensive = unique and unmatched by other UNIX or Linux operating systems

Example use:

  • smitty tcpip
  • Select the interface, add IP address + hostname + gateway + DNS? Hit the Return key and the job done
  • On the network in 6 seconds

More Information:


image

Nigel's Performance Monitor: 3rd place

Briefly:

  • Instantly, on-screen find how AIX is running
  • Or save stats to a file for later processing or graphing

Benefits:

  • On-screen watch all the important stats: CPU, RAM, disk, network & loads more
  • Analyze the stats for graphing: nmon Analyzer, nmon Visualizer, pGraph & nmonchart + more
  • nmon - Takes few CPU cycles to get lots of data
  • By popular demand its part of AIX now

Example use:

  • nmon Start the command with no options to enter interactive mode. Then, use this key to switch on and off displaying the performance stats:
    c=CPU
    m=memory
    n=network
    d=disk graph
    t=top processes
    h=help
  • Capture the statistics every 5 minutes all-day 
    • nmon -fT -s 300 -c 288
  • Graph the statistics 
    • nmonchart   hostname_datetime.nmon   hostname_datetime.html
      Then open the html file with a browser

More Information:


image

Live Partition Mobility (LPM): joint 4th place

Briefly:

  • Jumping a virtual machine between two physical servers with no downtime

Benefits:

  • Greatly increased sys-admin flexibility
  • Load balance workloads across a 'cluster' of servers to maximum performance
  • Evacuate a Server for maintenance or upgrade
  • Use a new server on day one to maximize investment returns

Example use:

  • Move virtual machines without the user or application owners permission
  • Move VM to faster servers during critical periods

More Information:


image

Dynamic Logical Partition (DLPAR): joint 4th place

Briefly:

  • Change the number and settings of CPUs, the size of memory & add or remove adapters without a shutdown

Benefits:

  • Without stopping the running AIX applications:
    • Add or reduce CPU resources: VP or Entitlement
    • Add or reduce the amount of memory
    • Add or remove virtual or physical adapters

Example use:

  • Grow the number of CPU or the size of the memory of an LPAR that needs drastically more performance in a peak
  • Shrink an LPAR in slack times or longer-term reduces workloads to reuse the resources for use elsewhere
  • Immediately react to changing workloads demands.

More Information:


image

Network Installation Manager (NIM): 6th place

Briefly:

  • Installing AIX over the network + updating AIX over a network

Benefits:

  • Fast and efficient installing of AIX from a single server
  • A single place to store install and update images
  • mksysb Backup Images - NIM can install these images for rapid total OS recovery

Example use:

  • NIM defines a virtual machine and image
  • Start the VM and point to NIM and it does the rest in ~10 minutes
  • Roll out service pack updates to many AIX virtual machines at the same time

More Information:


image

Make System Backup (mksysb): joint 7th place

Briefly:

  • Make a bootable image of the rootvg & save to a disk or on NIM

Benefits:

  • This function allows quick and complete recovery of a failed rootvg (the AIX operating system disks)
  • It can be used directly to disk or tape or used to make a CDROM
  • Also, the image can be used to network installation of AIX to rebuild a failed or corrupted root volume group.

Example use:

  • makesysb -i /backup/June2nd.mksysb
    
    # backup to tape and exclude files in etc/exclude.rootvg
    makesysb -e -i /dev/rmt1 
    
    # use a JFS2 snapshot
    makesysb -Ti /mydump/backup.mksysb 
More Information:


image

Logical Volume Manager (LVM): joint 7th place

Briefly:

  • Groups disks (physical volumes (PV)) into Volume Groups (VG) & then allow the creation of logical volumes (LV) of many types across the group

Benefits:

  • By grouping disks (or LUNs) LVM allows many advanced features for the virtual disks (logical volumes) like:
    • Spanning multiple disks,
    • Mirroring data for redundancy,
    • Striping for performance,
    • Migrations to or from disks,
  • Export or import the whole volume group between servers
  • Also, supported ad boot, jfs2log, sysdump, raw and paging logical volumes

Example use:

  • Volume groups:
    • lsvg # list the VG names
      lsvg rootvg # list the VG specification options
      lsvg -l rootvg # list of logical volumes
  • Logical volumes:
    • lslv hd4 # list the LV options;
      lslv -l hd4 # shows the LV layout [hd4 is the root file system]
  • Physical volumes:
    • lspv # list all the physical disks;
      lspv hdisk1 # disk details
      lspv -l hdisk1 # list lv’s

More Information:


image

PowerHA: Power High Availability 

Briefly:

  • Allows two (or more) AIX virtual machines with shared access to disk to back up each other for quick recovery

Benefits:

  • Once any service fails, the backup machine brings the service or application backup
  • An automated takeover in an emergency + manual failover for maintenance
  • Monitoring for many issues all the time: network, disks, application code, server HW, ...

Example use:

  • One to one backup across the room or campus or can be across large distances
  • Or configured in a group (like 1 backup server to 5 production servers)

More Information:


image

JFS2 - Journal File System 2

Briefly:

  • Enterprise-class file system with decades of field hardening

Benefits:

  • Fast and reliable access to data and files
  • JFS2 allows the live growing and shrinking the file system size including the root file system
  • Direct I/O & Concurrent I/O for RDBMS caching their own data
  • Snapshots allow rapid returning to know point-in-time and backups
  • Log (journal) of structural changes to remove rebuilds on server failure

Example use:

  • Full file system life-cycle control with a quick and simple smitty or command-line interface
  • There is only one excellent file system on AIX and that means avoiding having to decide between many good alternatives

More Information:


image

PowerVC: Power Virtualization Center

Briefly:

  • OpenStack based full virtual machine life cycle management with a "cool" GUI giving Enterprise Cloud for Power Systems adding top POWER features

Benefits:

  • Reduced system admin man-power for VM and application stack rollouts
  • Repeatability and consistent AIX images
  • OpenStack but with all the POWER and AIX benefits on top LPM, Dynamic LPAR, Storage space control, large-scale Cloud environments
  • Can include self-service by IT client groups

Example use:

  • Spin-off a new VM with AIX and database in seconds of a size that is simple to decide
  • Add extra disks or grow virtual disks in seconds

More Information:


image

Power Systems Enterprise Pools

Briefly:

  • By licensing CPU cores & Memory GB’s (instead of buying physics cores & RAM), you can move the active cores across a group of Enterprise servers at any time

Benefits:

  • Flexibility is where you run applications and their HA counterparts
  • Able to drain a server of application for service by moving the activations
  • Move CPU cores and memory to application temporarily needing more for peaks
  • More licenses between generations of hardware

Example use:

  • A six-pack of E880’s with a 768 CPU core pool & 24 TB of RAM? Then, you decide where to activate them today!

More Information:


image

Live Kernel Update (LKU)

Briefly:

  • Updating the AIX kernel without stopping the kernel

Benefits:

  • No longer requires an AIX reboot to activate a new kernel-level when you upgrade AIX to a new Technology Level or service pack updates.
  • No application outage
  • Internally, it uses WPAR technology to achieve it

Example use:

  • Get authenticated to the HMC:
    hmcauth -u hscroot -a hmc_name
  • Use the underlying command for AIX updates:
    geninstall -k -d /tmp efix.150813.epkg.Z
  • This update takes approximately 20 minutes.

More Information:


image

Scalability - large virtual machines

Briefly:

  • AIX is developed for large-size LPARs for decades

Benefits:

  • AIX runs happily on the largest Power Systems servers available
  • Up to 192 CPU Cores
  • With SMT=8 that is 1536 CPU Core Threads
  • Also, worked on the 256 CPU core POWER7 processor based Servers in the past
  • With the release of the Power10 servers in 2021, the maximum is 240 CPU cores and 8 times that in CPU Core Threads

Example use:

  • Large RDBMS there are examples of 192 CPU core LPARs running Oracle RDBMS
  • Other large workloads get to run on large servers and thus avoid the overheads of running across a cluster of servers

More Information:


image

System Administrator Stable Command Line Interface

Briefly:

  • Many decades without significant command-line syntax changes
  • Unlike Linux based operating systems

Benefits:

  • No man-power lost to relearning different commands
  • No command changes between AIX upgrades
  • Additions features are handled mainly by options

Example use:

  • smitty
  • The basic commands like ksh, vi, awk, sed, grep, sort
  • All the admin commands for LVM, JFS2, PowerVM, AME, NIM, mksysb features

More Information:

  • This feature is due to an AIX development policy of no System Admin surprises as you upgrade (don't change the defaults), constraint in not randomly changing features and thus annoying users, and "If it is not broken, don't fix it"
  • This advantage is only possible, if you get commands properly designed from the start.
  • So no further references.


image

Active Memory Expansion (AME)

Briefly:

  • Make memory seem larger than it really is by compressing memory

Benefits:

  • Improves the performance of the workload by storing many compressed memory pages in a cached area and then paging from there on demand
  • Either reduce the size of the LPAR = freeing up RAM for use elsewhere
  • Or make the LPAR seem like it has lots more memory

Example use:

  • amepat: This command predicts the effect of various RAM sizes being traded for CPU time compressing and decompressing
  • Switch on and set AMD Expansion target on the HMC
  • Monitor performance by running the commands topas or nmon or lparstat -c 1

More Information


image

Shared Storage Pools (SSP)

Briefly:

  • Simple to use VIOS distributed file system on top of LUNs

Benefits:

  • Subsecond allocate and assign virtual disks to virtual machines = reduced sys-admin time
  • Spreads I/O across LUNs = fast I/O even for single disk VM
  • Automatic mirror & resilver and tiers offer different disk type grouping
  • LPM ready by default and simple data migration between disk vendors
  • AIX now supported 64 KB page expansion.

Example use:

  • Cluster 12 dual VIOS servers for fast I/O, less admin time, and Live Partition Mobility (LPM)
  • Disk space managed by Server admin or the VIOS admin staff
  • Supports PowerVC regardless of underlying disk vendor

More Information


image

Binary Compatibility

Briefly:

  • Today's applications definitely run on next years AIX.
  • With a 30-year history of innovation, AIX continues to deliver on its robust roadmap with every release. The long-standing AIX focus on binary compatibility allows applications to run unchanged and without recompilation on the newest releases - guaranteed.
  • The Hexadecimal in the badge spells "AIX 3.0 to AIX 7.2" but you knew that right!

Benefits:

  • Applications from AIX 3.0 (~20 years ago) run happily on AIX 7.2 today
  • Reduced needs to upgrade applications to move to newer AIX versions
  • Hardware independence allows migration to new faster and less expensive HW and reduces software license

Example use:

  • No risk AIX upgrades for high performance, high-security OS
  • Stability of your code stack

More Information


image

Advanced 'O' Tuning Commands

Briefly:

  • A consistent set of commands to tune AIX in many areas

Benefits:

  1. asoo For tuning the Active System Optimizer
  2. ioo For tuning the I/O
  3. lvmo For tuning the LVM options
  4. nfs For tuning the Network File System
  5. no For network optimization
  6. raso For tuning the RAS features
  7. schedo For tuning the CPU Scheduler
  8. vmo For adjusting the virtual machines

Example use:

  • List the options, min, max, default & current values: vmo -L
  • Change current & reboot setting:
    ​​​​​​​schedo -p -o vpm_throughput_mode=2

More Information


image

Simplified Remote Restart (SRR)

Briefly:

  • If you shut down a Server or it halts due to a problem, then the LPARs (VMs) are automatically and in priority order be restarted on other servers in the group, if controlled by PowerVC.
  • Otherwise, you can run an HMC command to restart the LPARs and select a target server.

Benefits:

  • Simple to set up
  • Allows problem recovery with no user interaction and auto return to service
  • Always ready to catch a whole server fault
  • These functions require PowerVC to be controlling the Server & LPAR plus the LPAR has no physical adapters and resources (like when LPM ready)

Example use:

  • Use the HMC Enhanced+ GUI to set the SSR flag & check the state of SRR

More Information


image Note: The newer dnf replaces the older Yum command

AIX Toolbox for Open Source & Linux Applications

Briefly:

  • IBM makes these popular apps and tools available for AIX users

Benefits:

  • Trusted source of the commands
  • One place to look for this type of tools

Example use:

  • bash, curt, gcc, Git, gimp, gzip,
  • HTTPd (apache), MySQL,
  • perl, php, python, ruby,
  • samba, squid, sudo,
  • VNC, vim, wget, dnf
  • And hundreds more.

More Information


image

PowerSC: Power Security & Compliance

Briefly:

  • A toolset of many security features & 1 for added performance

Benefits:

  • This product comes with six separate features:
  1. Trusted Boot - Be sure that boot media
  2. Trusted Network Connect - on AIX startup & ensure a minimum AIX level
  3. Trusted Firewall - Pass packets between LPARs with no external firewall
  4. Trusted Logging - Secure audit files away and safe from malicious modification
  5. AIX Real-time alerts - Immediate action - no more periodic script running or polling
  6. Trusted Surveyor - Checks all LPARs on a VLAN + reports changes

Example use:

  • Too large a subject for this article due to the high number of functions in this product.

More Information:


image

AIX Workload Partitions (WPAR)

Briefly:

  • Separates processes and all connected features into a defined group, can be managed as a "mini" AIX with separate resources
  • This function is similar to "Linux Containers" years before they were thought up and with rigorous security.

Benefits:

  • On large AIX virtual machines, you can monitor the CPU, memory, and I/O at app level
  • Understand the resources used by different apps and control them
  • A WPAR can be moved between AIX images
  • Start and stop application separately

Example use:

  • Separate the RDBMS, web server, and backup for CPU monitoring & limiting
  • Separate WPARs or join them on one AIX at will

More Information


image

AIX Flash Cache or VIOS Flash Cache

Briefly:

  • Cache disk block I/O for faster disk access to a Flash or SSD device
  • Yields high I/O performance and reduced SAN traffic

Benefits:

  • Cache SAN-based disk I/O to local AIX or virtual VIOS solid-state disks
  • Massive improvements to high disk I/O workloads
  • Can be instantly switched off to provide Mobility to target with or without Flash cache

Example use:

  • RDBMS performance increased
  • Any other workloads with regular read activity benefits too.

More Information


image

aixpert: A command in AIX for operating system Security Hardening

Briefly:

  • Apply up to 470 rules to raise AIX security

Benefits:

  • Rules maintained by IBM, you can add your own or comment some out
  • Save your current rule set as a central Network Installation Manager (NIM) & apply it to hundreds of AIX images
  • Run reports to ensure the rules not altered or determine unexpected changes due to a hacker

Example use:

  • Change the root password then: aixpert -l high (lowercase L)
  • Check the current settings: aixpert -c
  • Capture the current settings: aixpert -l high -n -o /tmp/setting.xml

More Information


image

Role Based Access Control Security (RBAC)

Briefly:

  • Assign AIX users only the admin roles & system resource they need for their job

Benefits:

  • Fine control of specific actions on specific resource for particular users and user groups
  • Avoids root user access and so removes the risks in systems administration

Example use:
Check the AIX manual pages for:

  • mkrole, lsrole, chrole, rmrole
  • mkauth, lsauth, chauth, rmauth
  • setsecattr, lssecattr, chsecattr, secattr

More Information


image

AIX Active System Optimizer (ASO)

Briefly:

  • AIX Active Systems Optimizer constantly monitors & optimizes AIX for better performance
  • ASO includes the Dynamic Systems Optimizer (DSO)

Benefits:

  • Automatically tunes once started
  • Optimize CPU cache and memory for affinity
  • Optimize memory size pages for efficiency
  • Optimize memory pre-fetch

Example use:

  • startsrc -s aso
    asoo -op aso_active=1
    tail -f /var/log/aso/aso_process.log

More Information


image

First Failure Data Capture (FFDC)

Briefly:

  • On a system failure, the FFDC feature captures a complete system state & error reports

Benefits:

  • State is used to unconfigure on start-up actual components that failed based on complete information rather than guesswork
  • Avoids repeated restarts with fault components online
  • RAS is increased with less downtime and correct component replaces by CE

Example use:

  • Hopefully, you never see this problem!
  • If you have a failure the restart removes the correct "failed" components, first time and CE replace the components only the once.

More Information


image

Field Hardening

Briefly:

  • With many millions, AIX images running every day for 30+ years all that experience is used to make AIX tough to stop

Benefits:

  • Every AIX & HW generation is built on previous experience and features added to reduce outage and increase Reliability, Availability, and Serviceability (RAS)
  • You get the results on millions of uptime years and AIX coded to withstand many real-life problems in computer rooms around the world

Example use:

  • AIX handles all sorts of environmental issues - Clock down, use alternative paths
  • AIX works around failed computer parts - RAS built-in
  • AIX captures failure details to allow fast rebooting and resolving - FFDC

More Information

  • Field hardening is based on years of experience. IBM analyzes all problems on all servers at clients site and determines how the number of problems can be reduced by add features to
    1. Stop the problem happening,
    2. How to provide faster diagnostic,
    3. How to isolate or work around the problem live and then
    4. How to more simply recover from the problem.
  • This process reduces the number problems in following years.
  • It is a long-term commitment to improved Reliability, Availability, and Serviceability (RAS) every year.


image

Encrypted JFS2

Briefly:

  • A feature of JFS2 so it is available at any time

Benefits:

  • Encrypt at multiple levels: whole file system, directory, or a single file level
  • Can allow the root user to back up without the data being "in the clear"
  • Simple to operate and transparent to applications (if they have the keys)

Example use:

  • Make the entire database encrypted
  • Have a special directory area for your secret files
  • Fulfill your regulation security requirements with built-in functions

More Information:

More Information


image

Memory Protection Keys

Briefly:

  • Programming service to catch rogue pointer use and a larger source of bugs

Benefits:

  • Limits a program's access to certain memory regions
  • Used internally by the AIX Kernel to catch badly written device drivers
  • Used by RDBMS user-written functions within the RDBMS server

Example use:

  • Read the AIX manual pages about:
    • ukeyset_t C data structure
    • ukeyset_init() and ukeyset_activate()

More Information 


image

ProbeVue Debugger

Briefly:

  • Dynamic trace sessions of user code without modification

Benefits:

  • Dynamically add user-specified probes in code & attach actions to print data
  • Allows understanding of code, data & error conditions
  • Typically, monitoring functions or syscall entry and exit or regular intervals
  • Used probevue by using a script in "C" like syntax allowing others to reuse them
  • Low effect on performance

Example use:

  • probevue myscript.e

More Information


Additional Information


Other places to find Nigel Griffiths IBM (retired)


  •  

Document Location

Worldwide

[{"Line of Business":{"code":"LOB08","label":"Cognitive Systems"},"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SWG10","label":"AIX"},"ARM Category":[{"code":"","label":""}],"Platform":[{"code":"PF002","label":"AIX"}],"Version":"All Versions"}]

Document Information

Modified date:
03 July 2023

UID

ibm11115577