A thing AIX would do it like no other Linux distro?

Hi,

I would like to know, is there a thing that AIX would do it, and RHEL or SLES would not? Something specific and great in the same time.

It might sound weird, but I'm very curios.

Thanks a lot guys!

Non-serious answer: AIX will cost you money. :slight_smile:

Serious answer: My understanding of AIX is that it is (usually) sold with proprietary hardware. That means that device drivers are "guaranteed" to work, eliminating the driver incompatibilities and quirks of that sort with SuSe/RHEL (and other Unices, such as FreeBSD).

If you look for a really reliable, mature cluster solution you will have to consider HACMP, which in turn runs only on AIX.

Virtualisation is also a thing which is done in that level of maturity only in AIX and nowhere else. I guess it pays off having 40+ years of experience in building mainframes when it comes to implement virtualisation techniques.

I hope this helps.

bakunin

Thanks guys!

Well, Linux has it's own Linux-HA cluster sollution, but is there a real big advantage, something that only AIX has?

I don't know much about the other systems but AIX Virtualization (aka PowerVM) has a feature called Live Partition Mobility. It's the ability to move a complete LPAR from one hardware frame to another (as long as they are both on the same HMC) w/o any downtime for the end user.

i would say AIX's LVM (logical volume manager) and JFS2 are distinct to AIX. its solid and stable for disk management where solaris relies on veritas and linux uses ext3.

mksysb is another great feature of AIX. being able to take system images and store them onto a nim server makes system administration easier.

solaris uses ZFS which is unique and very easy to manage. hp-ux comes with veritas (if i remember right).

I would say:

the ODM - a database that keeps track about every piece of hardware, software, firmware, driver, setting or anything else implemented on the server - and that perfectly cooperates with Smit/Smitty/WSMIT - system administration interfaces that can help system administrators to avoid big mistakes ...

VMM - a subsystem that manages memory memory usage and allocation

WLM - workload manager that allows you to move memory cpu and IO between applications

LVM containing JFS2, and scalable Volumegroups that allow you to create filesystems and volumegroups in sizes that no other OS will allow you to - and that lets you create and drop volumegroups, logical volumes, filesystems and disks on the fly - but still allows you to run other volumemanagers like VXVM in parallel if your application does require it

SRC - system resource controller - that provides you with an interface to manage and control your subsystems without any further manual intervention

SWVPD - software vital product data - keeps your software on track, manages versions, history, installable options of your software packages

and regarding hardware: the unique opportunity to run AIX, Linux and Mainframe on the same hardware and the same subset of virtualization

Rgds
zxmaus