Following a bit of reading (turns out not enough), including an article on Cloud performance on the thebitsource ,  we concluded that our small scale development app which relies on a single MySQL server would be more than catered for by a RackSpace cloud server. We needed a small (but consistent) IO requirement and modest memory/CPU. 

The first month of service was a complete success and we began to consider migration of live systems to cloud servers; then we where suddenly hit by a dramatic IO performance drop lasting 6 hours making the instance unusable during the online day. 

IBM are gradually releasing the features from the next generation of POWER7 VIOS. In short VIO servers will be able to participate in a cluster outside of a single frame (across multiple boxes). This will mean that you can present virtual disks to LPAR's from one VIO server and make them available to another server via another VIO server at the same time without using NPIV. This reduces the dependency on feature rich storage (and cooperative storage teams!) will prove useful for the scale-out clustering technologies and for partition mobility.

They are planning to build in some useful auto-provisioning, Snapshot, cloning and thin provisioning features at the VIO cluster level all controller by director. Making it much more flexible to work in a virtualised environment with multiple systems.

If you are up to date with your VMware versions and you are using the latest in backup software (Netbackup 7, TSM6 veeam, vRanger etc) you will enjoying the pleasure of vmdk image backups without client software in each VM.

There are some good reasons for wanting to do VMware image backups:

I've been asked this question quite a lot lately, and to be honest I've been a bit stumped, other than a  gut feeling that monitoring from the OS as well as the VMware layer is a good idea, I had little else to offer the argument. So I figured I should take a more serious look at the monitoring available within VMware vCenter and come up with a more compelling argument. Perhaps the more seasoned VMware admins out their could comment to offer their thoughts on the topic?

vCenter for VMware provides an excellent selection of performance graphs enabling you to track the CPU, memory, network and I/O consumption for each VM as well as the ESX host itself. You can also set "alarms" that trigger e-mail notifications. So why would I ever need an additional agent based OS monitoring package? What more could you possibly need? Do the OS metrics become less important in a VM world?

( 3 Votes )

The new rPerf document from IBM released on the 17th of Aug contains some big rperf numbers for the IBM System P POWER7 795, and due to the dramatic leap in processor core count (64 on a 595, to 256 on a 795) they have paid some thought to the way they display the rPerf figures.

( 0 Votes )

I was interviewed the other day by an IT journalist who was posing some interesting questions on the effect of virtualisation on IT staff - do they need to have different skills? do we need more or less of them? usual kind of questions. But as we discussed the ramifications of increasing the level of virtualisation one area I was keen to press home was the need for really good communication and shared objectives between the IT staff that understand the technology and the accountants who count the cost.

 

On the hunt for some tools for managing IBM POWER boxes running multiple (AIX and linux ) LPAR's ? You will find this link useful:

http://www.ibm.com/developerworks/wikis/display/WikiPtype/Performance+Other+Tools

( 3 Votes )
Following a LinkedIn question posted on how virtualisation technologies work I wanted to share some information on the Why of virtualisation. So much of the virtualisation marketting is geared to the statement that it will save costs but as anyone who has looked at an IT budget breakdown the word cost in this context is far too vague to be of value. Its a bit like saying air is needed for breathing. OK yes it is in a broad context but actually oxygen is the important component (and even then it could be toxic under certain presurised environments so mix concentrations need to be selected based on usage)

We all know the pitch, consolidate your storage, virtualise your servers and share the infrastructure where possible  through use of a blade based solution for the best possible TCO.  Why would anyone do anything else? We are all agreed. OK lets look at some prices to back up our cost busting claims. I have used list prices throughout. Discounts will vary according to customer, vendor and the time of year.

Example 1

Customer has been buying HP-UX servers and storage for one application and IBM AIX servers and storage for another application that often sit together in the same rack. Would it make sense for both systems to use the same storage? Of course it will. OK what's is the price? The HP storage being used is an MSA60 SAS attached with 9 disks....

 

Whoopee! the feature we have all being waiting for. Active memory sharing.

This will mean that you can oversubscribe memory on a POWER6 system so that you let the system deploy memory where you need it the most whilst the LPAR believes it is always there (so no DLPAR operation required). It is the final piece of the jigsaw to creating a fully virtualised environment so that you can make the most out of consolidating different AIX (and linux) systems together.