Request For Topics

Hi everyone!

If you're a regular reader of this blog, or other EMC blogs, you know we can cover a lot of different topics -- and they don't all have to be storage related!

I thought I'd put a quick note out and ask my regular readers if there are any themes or topics that you'd like covered -- either new ones, or deeper dives in current themes?

Just to set expectations, I'm not saying I can deliver on what people want to hear more about, or that other EMC bloggers can do so, but I'd be really curious as to what other people are really curious about.

If you have a thought or two, would you mind sharing it in the comments section?

Thanks!

Comments

[Pascal]

Hi Chuck,

I heared your keynote at the EMC Forum 2009 last week in Frankfurt, but I was a little bit disappointed that you left so early. I really like the ideas about the internal clouds and I think my company is on the right way with EMC and VMware. But I see one "problem" and would like to hear your opinion to this:

In the past every some years the server hardware needed to be changed for some reasons: The next Symmetrix generation needed some updates or the hardware was to slow or something else. And this was mostly the time for OS updates, so old OS version died a "natural death". But within a VM pool or a cloud there is often no reason for the OS updates. Imho we will get more and more old OS instances within our pools/clouds that will become really bad to maintain.

I'm interested in your thougts about this.

Best regards,

Pascal

[Nick Mehta]

Hi Chuck,

I enjoy your blog and read it every time you post. Insightful stuff.

I'd be curious as to your thoughts on how migration to cloud-based apps (e.g., salesforce.com) changes customer approaches to concepts like data protection. Do they just trust the cloud provider to do it? Do they start making "backups" of their cloud-based data themselves? To another provider? To a local store?

Basically how does data protection change in a world when most of your data lives in the cloud?

Nick

[Chuck Hollis]

Hi Pascal

I'm sorry I wasn't available in Frankfurt, but they keep me pretty busy on these trips. In addition to the keynote session you saw, I had several other meeting pre-arranged.

You make an interesting point that I hadn't considered before, e.g. that there is a natural cycle of life and death regarding IT technology (specifically operating systems), and -- by potentially prolonging the life of an obsolete operating system via virtualization, do we eliminate a natural "predator" that takes out the old, and makes way for the new?

Indeed, I hear stories about people running old MS-DOS 6.2 code on virtual machines, so the technology will allow that sort of prolonged "life extension" in a way we never could have contemplated before.

On the whole, though, I think having choices are a good thing, and virtualization gives you more choices: stay with your existing version, or move to a new one. And, if you do go to a new one, the process is easier and less risky than before.

Whether people use these new choices to make poor decisions or not -- well, that's a different story.

Thanks for reading, and thanks for your question -- even though I don't think I have a good answer!

-- Chuck

[Chuck Hollis]

Hi Nick

I would offer that any decent cloud-like offering would offer you choices: backed up by the service provider, backed up to your site, or backed up to some other service provider.

In one of my recent posts, I attempted to delineate a stack of abstract/service layers, with the idea that any layer could be provided internally, or externally.

Backing up information (an information-related service) is below the app/bus process layer, so there should be no logical reason why those services shouldn't be considered separately with choices all around.

For example, I routinely back up this blog to one of my desktop devices. It's not that I don't trust the TypePad people :-)

The other issue is ownership of information. Imagine that a given app provider had a Really Bad Day, and you just couldn't get to your information for a prolonged amount of time (legal issues, bankruptcy, massive infrastructure failure, etc.)

The prudent IT consumer would most *definitely* want a usable copy of the information (if not the application itself!) in a separate location or service.

Great question -- I'll try to cover it in more detail in a future post.

-- Chuck

[Michae Teschner]

Chuck,

I am a fan of your block and I think this is extremely usefull.

I would be interested in topics that cover the benefit that comes from the combination of different EMC technologies and business units in perspective of customer / business problems. What we do that others can't.

e.g. the new aquisition of Kazeon is now part of CMA, but there could be also a benefit of the technology for DLP and thus having a consistent data classification for eDiscovery and information protection.

best regards
Michael

[John Kane]

Hi Chuck

I heard that when EMC bought RSA you were going to change your strap line to EMC - Where information lives securely - what ever happened to that!?

[shiningarts]

There is a traditional Korean saying that says when whales do battle, shrimp get hurt. While the saying no doubt holds true in many cases, EMC may prove to be the exception to that rule in the midst of the Intel structural shakeup (which is due, in part, to the fact that Moore’s law is about to smash into a brick wall). As former Intel executive Pat Gelsinger joins EMC as president and COO, it will be real interesting to see how this plays out.

Imagine, if you will, a chip directly on a storage device. If this doesn’t meet with the same ill-fated end as Intel’s attempt to add a chip to a communication device, it could change (dare I say “revolutionize?”) the landscape of storage architecture altogether. Think about what would happen to process IO, dedupe, archive, security, virtualization, or anything else if these processes happen right at the storage devices instead of from a central processor. Eventually, this would make cloud computing even better as far as storage is concerned. In addition, this would ultimately accelerate intrinsic object-oriented file architecture. Just a thought. . . .