OOW 2015, day 1 -- Essbase in the Cloud

It’s here, it’s here, it’s here

And it’s for real.  Essbase Cloud Service, aka EssCS has arrived.  And it’s pretty awesome.

I first heard about this from ER who found the release docs on Network54.  And then I promptly asked for his post to be deleted because I had no idea this was about to drop.  Oracle played the release pretty close to the vest.  Some partners may have been involved in a beta, but not me.  Or MMIC Glenn Schwartzberg, and if he didn’t know, I think very few do.

For details on all of the things that Essbase cloud means, I encourage you to read fellow ACE Director Eric Helmer’s excellent blog post.

I also encourage you to take a looky-look at the Essbase 12c documentation.

I’m not going to restate what either document, but instead give you an experiential take on what it’s like to use it from a developer’s point of view.  Yes, this is Cameron the touchy-feely guy.

A note about my mad photography skilz

The screenshots are, for the most part, taken from the side of the demo screen.  I wanted to stick my big head in front of everyone to take these snaps but thought I’d be torn limb from limb by the ultra Essbase geeks in attendance.  Also, I can’t take pictures to save my life.

With that caveat, off we go.

At the demo grounds with my peeps

Here’s Kumar Ramaiyer, Essbase development manager, and Natalie Delemar, ODTUG Kscope chairwoman, ODTUG Vice President, and all-around Essbase geek, about to go into the demo.  The crowd grew and grew.

More and more

The gaggle of geeks got bigger and bigger.  There’s the back of Gary Crisci’s head as well as the Planning product manager Shankar Viswanathan.  

The men responsible

Steve Liebermensch, where are you?  In his absence, here’s Gabby Rubin and Kumar Ramaiyer, product and development management.

The obligatory Safe Harbor

It wouldn’t be a demo without Oracle’s warning that what you see may indeed come true, or may never happen, and you shouldn’t make any decisions other than giving in to an extreme Essbase geekout.  Seriously, don’t drive your company’s strategy, commitments, etc., etc., etc. through what you see here (or in the demo grounds) because it could all change.

You’re looking at alpha code and I suspect many of the limitations will go away by the time it ships but who really knows.

Logging into the cloud

Where’s EAS?  Where’s Studio?  Nowhere to be found in the cloud.  They’re just too chatty.  But they will be available on-premises, at least for a while.  The strategy is evolving but I think they aren’t going away any time soon.

One note about the database you’re about to see:  Hybrid is the standard, which I find fascinating, although ASO can be built in a different way.

Importing a database into the cloud from metadata in Excel, yes, Excel

I’ve likely bored you with my first experience with Essbase:  1993 (or 1994 as I no longer remember), OS/2, Essbase 3.1, a Compaq 66 MHz 256 MB server under my desk.  Essbase was largely a departmental solution, managed by finance super users.  

That was then, this is now, and now Essbase is an enterprise level tool with all of the infrastructure complexity that implies.  

It isn’t that EssCS won’t be suitable for the enterprise, it’s that it will also be suitable for smaller instances.  This will, I think, lead to an Essbase renaissance as those enterprise chains will be broken.

Did you notice the >8 character name?  Finally, finally, finally, although my sort-of expertise of scrunching down a app/db name down to 8 characters will become obsolete with which I’m more than happy to see go bye-bye.

Yes, it’s Sample.Basic metadata in Excel

What does that mean?  It means users/admins can build Essbase databases not through a hierarchy management tool, e.g., EAS or Studio, but directly from metadata stored in Excel which is then uploaded.  

Excel is a tool almost everyone in every department uses and understands.  So why not have the analysis tool be the Essbase administration tool?  Exactly.

Do you see the tabs?  Each one is a separate dimension.  Within a given dimension, you can see how the metadata is defined as well as the parent-child hierarchy.  I have to confess I didn’t ask why all of the dimensions are defined here as well as in the separate tools.  I’ll try to find out.

Editing the outline in the admin console

Excel isn’t the only interface – you can still directly edit the dimension although not in a manner familiar to EAS/AppMan users.

Currency conversion is not dead

The 12c ReadMe says it’s no longer supported.  And yet here it is.  Is it different?  Good question but I have no idea.

Member formulas and Kumar’s finger

Member formulas can still be entered manually.

Auditing

Want a change log?  Here it is.

You’ve imported it, now export

Outlines go both ways – no different than EAS today.

Note the tabs, note the formula

We saw before that member formulas can be made directly in the admin console; they can also be done in the Excel metadata file.

Solve order?  BSO/Hybrid?

Where did that come from?  Shades of the ASO engine on top of BSO in Hybrid?  Probably.

Note well

Database notes are alive and well, as are data and metadata uploads.

Btw, love the >8 character names.  Like a rainstorm in a drought.

Run it and monitor it

There’s a console, conceptually similar to Planning, FDMEE, and Shared Services.

Load data

File, SQL, file name, abort on error, but NO LOAD RULES.  Actually they will be there, somehow, maybe on-premises only, but I’m not sure.  Fingers crossed on their timely demise but dreams don’t always come true.  :)

What’s going on?

Here’s the data load showing that all is well.

Did it work?

Attributes?  Yep.

There was a question today on Network54 re attributes in EssCS – it sure looks like it’s included.

Calc script editor

They’re not dead!  How could they be?

Sandboxing in Essbase

This was demoed at Kscope15, but it’s the concept of different users entering different values.  An admin looks at them via process management and decides what to merge/release.  It is Pretty Cool.

A list of scenarios.

Sandboxing process management

Commenting on a sandbox scenario

Scenario console

Loading data and metadata simultaneously to ASO

ASO isn’t dead.  In fact it’s the default for a combined data and metadata load.  Again through Excel although I believe .csv and database loads will also be possible, but don’t hold me or Oracle to that.

Can you mobile adminster EssCS?  Yep.

From the sublime to the ridiculous but not really.

First off, this is mega cool.

Secondly, I can actually see a use case on something a bit larger than on an iPhone 5 being generated on the fly – you’re in a meeting, need to look at results quickly, and aren’t in front of your laptop.  While today that scenario does require a laptop it means you popping off to your office to do just that.  It’s no longer necessary.

Here’s a load from DropBox.  Yes, DropBox.  OMG so cool.  And flexible.

Can it be edited on mobile?  Why yes it can.  

Smart View in the cloud

That’s not actually correct, this is on-premises Excel connecting to EssCS.

What does this all mean?

Here’s how I see it:

1. Essbase will – optionally – go back to its departmental roots.  Will it replace Exalytics?  No, it’s not there, and will likely never be.
2. Essbase applications will be easier to implement than ever.  
3. Does anyone love owning EPM architecture?  Just what I thought.  This obviates that ownership.  Cue general joy except for infrastructure consultants.
4. EssCS will lead, on-premises will follow.  It’s frustrating but like any other business, the Essbase team have to follow the money which in this case is Cloud funding.  There’s a commitment to bring functionality to on-premises but EssCS will always lead.
5. Given that EssCS is by default Hybrid, we have yet another confirmation that Hybrid is the direction for Essbase.  Cue general joy again.
6. Given Hybrid’s central role, upper level cross dimensional tuples will have to be supported.

There’s an awful lot I don’t know, and an even more awful lot I didn’t quite understand, and alas quite a bit I’ve already forgotten.  

What’s really importing is that Essbase is progressing, a lot, quickly.  We’re all going to be challenged to keep up with the tool.

It’s very exciting times.

Be seeing you.

Day -1 at OpenWorld 2015

Why -1?

The formal start of OpenWorld is tomorrow, but we (or at least I) know know that the User Group Sunday is the real start of this insanity known as Oracle’s World.

And yr. obt. svt. will continue his well-known insanity by trying to live blog this conference.  It’s a task I’ll fail at given the scope, but I’ll give it the old college try.

Session the second

I actually missed the first session because I was having breakfast with my Bestie Natalie Delemar.  I’ve sinned and am trying to make up for it through diligence and penance.  

And that brings me to Wayne Van Sluys’ session on SmartView and OBIEE.  No, Wayne is not part of my penance – surely that involves a hair shirt, gnashing of teeth, and ashes thrown upon my head – but instead of my diligence that I will apply to today henceforth.

As far as the two tools, I blow hot and cold.  Do I see the logic in using SmartView to access federated OBIEE data?  Yes.  Do I see this combination as being either as intuitive and simple as SmartView and Essbase?  No, but it’s doing a lot more.

If you want OBIEE in Excel, absolutely this is the way to go.  If you’re accessing Essbase (or Planning or HSF or HFM or almost all of the EPM suite), in my opinion it’s not a great fit.

Session the second

Edward Roske, my long-ago employer, is speaking on Planning outside of the box.

Some people see Edward the man, others see Edward the CEO, others see Edward the ACE Director, but I see Edward the Hat:

Here he is for real:

As always, Edward is both amusing and informative.

Session the third

This is where yr. obt. svt. bores/delights/stuns-with-his-obviously-wrong-insights-and-tips.  Okay, it wasn’t half-bad, and the room…the room was half full.

Session the fourth

I’m listening to Ron Moore call his approach to Essbase calculations “simple stupid”.  I’ve known Ron for a super long time and there isn’t that guy does that isn’t smart, smart, smart, so I think he wins today’s record for self-deprecation.

All kidding aside, my philosophy is keep it as simple as possible until I can’t.  Simple is hard enough; unbelievably clever is either impossible, so close to impossible that it isn’t worth doing, or is completely unsupportable.

Session the fifth

Sarah Zumbrum, fellow ODTUG board member, is giving her session on converting BSO to ASO tips:

Session the sixth and the last

Dave Collins and Mike Nader are presenting on OBIEE and Essbase.  What’s beyond super-cool is that the content is based on Mike’s chapter Developing Essbase Applications:  Hybrid Techniques and Practices.

Keep on checking

Having just slit my Oracle ACE Director throat, I encourage you to check back every hour or so to:

1. Watch me self-immolate
2. Get my not-entirely-informed opinion on the sessions
3. Experience the closest thing to actually being here

I’ll be updating every hour or so.

Be seeing you.

Addendum

The sessions are over, although not the rest of the day.  There’s a BI meetup to attend, a credit card to retrieve that I apparently left with a waitress (first time ever – either I am slipping in my dotage or OOW is that kind of conference or both), and then OTN’s ACE dinner.

Be seeing you tomorrow.

Traveling By Tube (And We Don't Mean In London)

Will the train of the future be a high-speed tube, not a railroad?  That’s inventor and entrepreneur Elon Musk’s and others’vision.  And Musk, the man who brought us the Tesla (all-electric car) and SpaceX (for-profit space rocket company) is putting his own money behind a proof-of-concept project for what he calls Hyperloop. 

The concept sound simple:  move passengers in a sealed tube through a series of giant pipes propelled by air pressure at speeds up to 700+ mph.  That would mean a trip from New York to DC would take 20 minutes.

But this is not a new concept.  In fact, the first experimental “subway” in New York City, Alfred Beach’s “Pneumatic Transit” proved back in 1870 that it would work.  Despite political opposition Beach secretly built a 300-foot-long subway under Broadway near City Hall, offering daring passengers a round-trip ride in the system’s only railcar, pushed and pulled

Beach Pneumatic - 1870

by air.  The system ran for almost 3 years and carried over 400,000 riders, 11,000 alone in the first two weeks.  The fare was 25 cents (equivalent to $18 today). Competing elevated railroad owners eventually won the City’s franchise and Beach’s system was abandoned.

Even Beach’s idea wasn’t new, as vast underground pneumatic tube systems in Paris and London were already delivering telegrams and mail by the 1850’s.  As recently as the 1960’s, office buildings in major cities were designed with pneumatic tube systems for inter-office mail.  Many banks still use pneumatic tubes at drive-up windows.

Hurtling through a tube may be fine for mail, but what about humans?  As a recent article in Smithsonian Magazine points out, the psychological factor of being enclosed in a sealed tube, traveling 700+ mph, is not that much different than flying in a jet…  maybe just a bit more claustrophobic.

Whether by train or plane, I always like to look out the window.  Seeing where we’re going is half the fun, even on a familiar route.  But wrapped in a metal tube inside a giant pipe afford no views at all.  Riding 31 miles in the Chunnel under the English Channel takes 20 minutes at today’s speeds, and that’s more than enough time for me, thank you very much.

Of greater concern are the propulsion methods and the sheer physics of accelerating and braking from near-supersonic speeds.  But the biggest challenge of all would be where to

Another Hyperloop Rendering

locate the “pipes” and how to acquire necessary land.

Like high-speed rail, it would make no sense to follow the median on Interstate 95 or the Metro-North / Amtrak rights of way with all their twists and turns.  And anyone crazy enough to invest in any project along the coastline with the inevitability of rising sea levels should probably think pontoons, not pipes.

It will be interesting to see if Musk’s and others’ Hyperloop concepts get off the ground (pun intended), but I don’t expect to ride such a system any distance in my lifetime.