Storagezilla

FCoE: Hooking you up.

In some previous posts on FCoE we introduced the EMC FCoE Tech Book and covered the fact that currently you can't route FCoE traffic through intermediate switches. Usually the next comment that comes up is "Okay, I'll just go direct connect for now and that'll give me an end to end solution."

No it won't.

You might think things work that way but they don't. Just as I discovered when I first looked into this.

Currently there's no support for Fibre Channel Name Server in the adapter cards, the FC Name Server used to pull attributes from the devices and is used in zoning operations.

Where's the Name Server as things stand today? 

In an FCoE switch. 

So for compliance with the current standard you end up deploying a switch even if your deployment topology could be served by direct connections.

Direct connect, like routing between intermediate switches and an Open Source driver is still off in the future.

We'll get there but we're not there yet.

---

You have been reading "FCoE: Hooking you up" by Storagezilla

This post is part of a series on the development state of FCoE at time of writing. You may also be interested in reading

FCoE: Idiots to the left of me, liars to the right. Stuck in the middle with users.

&

FCoe: Where we are now.

FCoE: Where we are now.

Following on from my previous post highlighting the EMC FCoE Tech Book, amongst the blizzard of nonsense press releases we see put out on FCoE sometimes the same one a few months apart, Stu put's his finger on the hot button issue around end to end FCoE deployment today.

 Note that configurations of having FCoE traffic go through multiple FCoE (Ethernet) switches will require the updates which are being worked on in FC-BB-6 (although a small expansion of configurations specifically with blade servers should be able to be supported soon).  FCoE today is a consolidation at the server and access layer – full end-to-end solutions with larger aggregation will take time.

Brass tacks is that routing FCoE traffic through intermediate switches is still a work in progress and a significant deployment limitation for end to end FCoE, that SAN topology you're used to having just not existing yet. Don't forget that when people crank up the hype train and try and sell you on tech which is still a work in progress. You can reap the benefits of FCoE in your existing FC environment today while keeping the deployment capabilities of your FC Fabric, but anyone promising you an end to end FCoE solution today is trying to sell you on taking a big step backwards.

---

You have been reading "FCoE: Where we are now" by Storagezilla

This post is part of a series on the development state of FCoE at time of writing. You may also be interested in reading

FCoE: Idiots to the left of me, liars to the right. Stuck in the middle with users.

&

FCoE: Hooking you up.

FCoE: Idiots to the left of me, liars to the right. Stuck in the middle with users.

Nigel has a nice FCoE post up. It makes one think that with the chatter in recent days and the repeated questions I'm getting it's clear that there's an entire bunch of vendor flapheads out there who either don't understand the technology correctly or are intentionally obscuring it's limititations in order to make a sale today with the hope if they stall long enough the technology will catch up to their hype much much later.

Idiots to the left, liars to the right.

On top of Nigel's post I'm linking to the EMC FCoE Techbook which breaks down what FCoE is and then goes up to it's eyeballs in topologies concerns and configurations with regard to how you can start using FCoE today without having a big gap where the idiots or the liars have fudged the tech.

It's regularly updated as new technology comes on stream and FCoE advances so keep an eye on it.

---

You have been reading "FCoE: Idiots to the left of me, liars to the right. Stuck in the middle with users" by Storagezilla

This post is part of a series on the development state of FCoE at time of writing. You may also be interested in reading

FCoE: Where we are now.

&

FCoE: Hooking you up.

The Great IT Skills Drought

Deloitte's Talent Retention report is out and skimming it makes for grisly reading. 49% of surveyed respondents are considering leaving their current job and on the opposite side of the table surveyed employers feel that a talent drought is crippling their business.

Is anyone surprised? 

On one hand employees are getting hammered with the do more with less platitude while on the other hand employers now have a skilled workforce base which is stretched thin and not up to speed.

I have two comments. When you do more with less first you cut fat, then muscle and then bone. Employers need to remember that in their attempts to boost the bottom line by taking as much out of IT as they can. The fat went first and now you're just mangling things. 

Sometimes it ends up that IT is farmed out in it's entirety to an outsourcer. Oh the regrets I've heard voiced from companies a few years after going down that path, but I'm sure it's worked out for some people.

Employees on the other hand are responsible for their own destiny and their own career. If you feel you're struggling with something you need to make it a personal objective to get up to speed. If you want courses, training and certification then push for them.

As an employee you are as valuable as you make yourself and worth not a penny more.

VMware customers drop the hammer

If you knew the number of turd-o-grams I get anytime I don't 100% agree with a partner in public in my place you'd have given up blogging years ago.

And yet here we are none the less.

I think what has me laughing most at Charles Phillips fig leaf of a response scantily covering Oracle's dignity is he just won't come out and say.. 

"We're building an Oracle only Mainframe. Deal with it."

I'd find that perfectly acceptable and it's not a case of me having to like or dislike it. 

Customers have come me and said (and I'm paraphrasing) "When are you going to support something that would be a bad idea for EMC's core business and do nothing but help EMC competitors?" and I simply tell them that's not going to happen.

I don't invent some poor excuse which mangles the technology involved and then try and change the subject. That's insulting to all parties. So I just say no. But if there's enough pressure applied by customers no can become yes. In some cases it already has become yes but only because there was a compelling economic case for it to do so.

Ultimately companies are in the business of serving the needs of their customers before their competitor does. If overnight it was more profitable to support some configuration or new technological direction and if this was a complete 180 from what was said yesterday I'd rather the change happen and the money be followed then get all religious around technological sacred cows or unprofitable thinking.

In vendor terms the difference between No and Yes is usually the perception of a bunch of money.

$13 Million in cost avoidance. $10 Million in cost savings.

EMC IT has goals the same way any other IT department does for all but EMCs largest customers their goals are probably larger than most. 

In the September edition of the EMC IT Virtualization Journey IT Director Paul Divittorio talks about his goals. $13M in cost avoidance and $10M in cost savings over the next five years. 

No marketing fluff, just hard targets and their approach to making those targets.

 

They'll succeed and meet their goals or they'll fail to meet those goals. But they're going to do it in front of you.

Are Point in Time copies backups?

Back to fundamentals again for a moment.

Yes point in time copies are backups but they don't protect you from a system failure. 

Getting a copy of the data off the system protects you from a system failure.

We've had occurrences of this show up in public already this year where an upgrade to a storage system resulted in the loss of millions of user photos. There is no number of snapshots that was going to bring that data back. The fact the data had been backed up off of the system is what allowed that data to be recovered and normal operation resumed.

I'm a big believer in rolling the last point in time copy of the day off the system and over to different media during the nightly backup window. Take some copies during the day, once every 8 hours isn't bad assuming it meets your RPO..

Recovery Point Objective (RPO)

[Data Recovery] The maximum acceptable time period prior to a failure or disaster during which changes to data may be lost as a consequence of recovery.

Data changes preceding the failure or disaster by at least this time period are preserved by recovery. Zero is a valid value and is equivalent to a "zero data loss" requirement. 

..and then have a backup application roll the final point in time copy of the day over to disk, virtual tape, physical tape or whatever backup media you're using so long as it isn't located on the same system. 

While point in time copies are backups the question you have is if you're only doing half the job when it comes to ensuring data recovery.

Pat Gelsinger to run Information Infrastructure Products at EMC

Pat "There's no speed but ramming speed" Gelsinger has been hired to oversee what amounts to a large part of the company. I haven't seen a respin of the org chart yet but I can pretty much guess who's reporting to whom now.

--Updated--

Now I have. RSA, CMA, Ionix and ESD will all report up to Pat as part of a combined Information Infrastructure Products Group. Frank Hauck will continue to lead ESD, Mark Lewis CMA, Art Coviello RSA and so on. 

Gelsinger is an interesting choice, I've seen him speak frequently and used to have a copy of his book Programming the 80386 when I was in college. I'll admit to having not read his subsequent books on balancing faith, family and employment though I know he published them.

The Juggling Act should be available on Kindle for anyone who does eBooks.

A different era requires lots of different thinking and storage is going through a transformation, moving out of the glass room and into every part of your life. Be that in your home or running under hyperscale applications.

Business as usual thinking need not apply.

--Updated--

Storagebod thinks I should take the advice in Pat's book. This is a point I'll concede and I'll commit to now adding it to my reading pile. After all, that pile is only two feet deep at the moment...

Welcome to VirtualizationVille

Everyone likes free stuff and with that in mind EMC Ionix are giving away VMware Compliance Checker for ESX over at VirtualizationVille.

Compliance Checker is a free, downloadable tool that provides a real time compliance check for multiple VMware ESX host servers at a time. This product is a robust utility providing detailed compliance checks against both the VMware Hardening Guidelines and the CIS benchmarks for ESX. Unlike other free tools in the market, Compliance Checker is a fully functional product. For example, you can print the reports that Compliance Checker produces and can run compliance checks across multiple ESX servers at once.

Security can be a nightmare for physical hosts, virtualisation is a multiplier so carry the compliance checking forward.

You can also crunch some numbers with their ROI tool and check out the Ionix solutions site.

VMworld wrap up from someone who wasn't there.

I wasn't at VMworld and therefore don't have the sense of euphoria most of the attendees appear to have but I do have some observations from watching streamed keynotes, reading tweets (Yes Twitter was the front line info distribution network for me this year) and browsing the press coverage. 

While there were comments about the amount of partner demos this year and I'll admit to breaking out the Dyson and getting some vacuuming done during one or two of the more particularly boring ones while watching the replays after work, this strikes me as a sign that we've entered the build phase of the Private Cloud journey.

The vision thing is always more fun to think about than the plucking, cleaning and stuffing of a bird after you've shot it but before it goes into the oven. Where we are right now is that feathers are flying and quite soon your hand is going where no man's hand has gone before.

The real work beyond the white board has only just begun and there's no artistry in this part. 

There is when you think about it and after you present it, assuming you did a good job it didn't turn black and you didn't need a fire extinguisher, but right now the process is  a bit unpleasant and parts of it smell funky.

So last year we did the vision thing. This year it's about how the nuts and bolts come together. Next year we should have a lot of people ready to show their work off to their peers.

To paraphrase Henry Ford that's assuming that end users actually want an automobile and not just a faster horse. For all the lip service paid to being open to innovation and new ideas data center folks still treat IT like they were making toothpaste.

Innovation goes as far as adding a minty stripe to the same old formula they've been using for donkey's years.

This isn't a bad thing when your job involves offering services, protecting information and making it available to end users so the reticence is understandable.