What is the single most important thing Oracle VM is missing?

I have been going through the Oracle VM feature set.

As a virtualization solution it actually looks pretty good. The cost is in the lower end and it seems pretty feature complete.

But there is one thing missing…one huge feature.

VM snapshot based backups. I’m betting that if they would throw a API for taking snapshot based backups and make a deal with Veeam to support it this would actually make quite a lot of system administrators take a harder look at using Oracle VM for at least some projects (i.e, virtualizing Oracle applications).

From my standpoint this is my biggest issue. Oracle has already gotten servers from multiple vendors certified (hcl) and it seems that Oracle is playing nicely with the larger hardware vendors (IBM/Lenovo, HP, Cisco).

Oracle – your techs might want to take a hard look at this feature – this will actually help you guys gain larger market share in data center virtualization!

Just my two cents!


3PAR StoreServ 7000 – Peer Persistence links

Howdy all,

I have been testing a 3PAR Peer Persistence setup using two 3PAR StoreServ 7200c, dual interconnected fabrics between sites and a multi-site VMware cluster (although only one node per site).

It works flawlessly!

Being able to take a array offline (disruptively, removing power to the controller shelf) and the only thing that happens is about ~10 second “delay” (while Peer Persistence fails the VMFS volumes over) for the virtual machines is pretty awesome.

We are still missing VVOL support for replicated volumes (and vMSC) but hopefully it will come later this year.

Here are the most important links on Peer Persistence:
VMware.com (KB article 2055904)
HP’s own Implementing vSphere Metro Storage Cluster using HP 3PAR Peer Persistence

And for a added bonus – HP 3PAR SSMC 2.1 makes Peer Persistence configuration easy as 1-2-3 by Techazine.com

The whole thing takes about 1 hour to configure when you know what you are doing and adding a new volume to the Peer Persistence configuration is a snap.


3PAR StoreServ 7000 – Zoning best practices

Howdy all,

Recently I started working with HP 3PAR StoreServ 7000 series storage. While installing them I was looking for the best-practice zoning information and seemed to find conflicting information. But after reading this document you see that HP recommends either:

  1. One initiator to one target
  2. One initiator to multiple targets

Check out the section called “FC hosts zoning” and things should be pretty clear! 🙂


A whitepaper on IBM’s HyperSwap


Checkout this whitepaper on IBM’s HyperSwap. The product is somewhat 1.0 and a little crippled (from my point of view anyway) but a large giant told me that most of the complicated stuff will be fixed in next version……


HP OneView – Quick overview


If you are looking for some information on HP OneView I recommend that you take a look at this site.

It is amazing how fast the development of OneView is progressing. I started using it back in January when we installed two HP C7000’s and 13 blades. Since then I have met a VP from the group which owns the OneView product inside HP, which was very clear on where HP was taking this product and how much HP is listening to the customers regarding features and development.

Now that version 2.0 is about to be released (FAQ here) I am really excited to see the new features, especially the features around storage (3PAR and the SAN). Having a single pane of glass to manage and monitor your whole infrastructure is great. The only “bad” thing is that of course OneView is mostly bound to HP products. It can monitor some third-party switches but with limited feature set.

Now, I guess we will have to wait for few more weeks for 2.0! 🙂


Disclaimer: I do not work for HP or a HP Partner. My views here are my own and have nothing to do with my employer!

Interesting announcement from IBM on Storwize/SVC software


If you are a user of the IBM San Volume Controller and/or the Storwize you should really checkout this announcement.

What I found most interesting was HyperSwap – now you can present a single volume from two different storage controllers which use metro mirror for a two way replication (although I do not have any more details on this). This is something like the Active-Active function in EMC’s Metro VPLEX installations. Cool stuff!


Testing Windows 10


I recently started using a “old” (4 year old) desktop tower to play Diablo 3 in co-op with my son. Last weekend I dug it up and installed Windows 10 on a spare 256GB SSD I had.

Damn! I must say they have actually out-done what they did when going from XP to Vista (I have to mention, I am probably one of the ten people in the world who actually liked Windows Vista:)). Everything is so smooth, even on hardware this old.

The new start menu is what the start menu should have been in Windows 8.1. However, after using Windows 10 I guess I have to forgive them….can’t wait until I can deploy this on my work laptop.

I have yet to use the browser they ship with it though….my Chrome habit is not something I see my self getting away from anytime soon!

Anyway….I must say, if you have a extra machine lying around (or some resources to run a VM), I suggest that you go grab the Windows 10 and give it a test run.


Yet another hardware refresh project comes (and goes!)

Howdy all,

I just finished a plan for redesigning/rebuilding one of the environments at work.

It is truly amazing how much time goes in designing, getting specifications, reading up on technical material, talking to industry experts and getting quotes from our server vendor.

There are more then 6 months since I started this planning and design phase with my colleagues. Our vendor has probably lost count of how many quotes I have requested due to this and I am amazed they actually still pick up the phone when I call them 🙂

What I am getting at is that often when we, the technical staff, start on a redesign projects small or large we are quite often stuck thinking about how awesome the newest CPU chips are or how fast some new memory tech is. When working on hardware refresh projects we are often replacing hardware that is three generations behind the new stuff. Speed is most often not the issue. Everything today revolves around software. We have to design around software, be it licenses or some other software flaw (or feature!).

After going through yet another project like this one might wonder – while we still have on-premise servers, would it be better to just “rent” hardware resources?

A thought for us all….