Archive for November, 2009

Offline for a few weeks…

November 18, 2009

Yes i know, it’s been a while since my last post anyway :P.

In a few days I’ll be heading back to my hometown of Sydney (via Tokyo – I love Japan) for a few weeks, so the internets will be a little less ranty for a little while. And unfortunately, I’ll be missing the next London VMUG which is on November 24th 😦

I’m looking forward to catching up with a bunch of people, including me ol’ china Jeff of Cloud Pimps. If I can convince him to get some green on his site, who knows what may happen to vinternals in the new year…

See y’all in a few weeks!

ECV – Utility Emerging

November 4, 2009

Todays announcement from EMC, Cisco and VMware (ECV) amounted to a declaration of war on the datacenter players in the industry as we know it. But it wasn’t the only thing that it amounted to – I see it as the first true sign of utility computing, and if the rest of the industry just sits back and takes this then it deserves to die.

Word on the street is that nobody wants to own anything anymore. CAPEX has become a dirty word. And large corporates have realised the hidden costs associated with managing many vendors – pricing, contract negotiation, relationship… these things have an often unseen cost for Enterprises in terms of human capital. The cost of designing, engineering and supporting hardware and software from many vendors was realised long ago, resulting in Enterprise standards based on the lowest common denominator.

The problem with this of course is that you end up with a solution that caters fairly well to 80% of the environment, isn’t optimal for any particular part of it and entirely misses 20%. And I’m not criticising that – for large Enterprises, it’s the only way to keep a reasonable grip on things. That is, if said Enterprises need to own the kit in their datacenters. But do they?

Virtual Hardware 7 Undocumented Feature Revealed!

November 2, 2009

So after my last post, VMware did some digging and the real reason for the initially odd behaviour was exposed – it’s a combination of the way Virtual Hardware 7 presents virtual disks to Guests, and the difference in default policy settings of Windows Server 2008 Standard Edition and Windows Server 2008 Enterprise Edition.

So in a nutshell, as Rob elaborates in the comments to my earlier post, it turns out Virtual Hardware 7 presents virtual disks as SAN devices – not as local disks. This resulted in the default SAN policy of WS2008EE offlining everything bar the boot volume during installation. To be honest, this is actually a good setting to have IMHO – I wouldnt trust Windows admins with clustering either. To anyone who I work with or have worked with, that last sentence was clearly a joke :D!

The VMware KB article has been drastically overhauled so no one else will be in the dark in the future, and hopefully stuff like this will also show up in the Guest OS installation guide where it really belongs.

So in the end we all learned something which is always good, and it goes to show that VMware’s response to potential bugs is in fact anything but woeful. Hurray!