THE SQL Server Blog Spot on the Web

Welcome to SQLblog.com - The SQL Server blog spot on the web Sign in | |
in Search

Denis Gobo

SQL Server In The Cloud Vaporware Or Inevitable?

After Paul Nielsen post SQL in the Cloud yesterday I decided to put on my Nostradamus hat and give you my take on this.

What needs to happen before we can move to the cloud?
There are a couple of things, here is a partial list


Performance has to be good
If performance is bad then nobody will move to the cloud. Since the data is external you probably want to have a secure connection and thus you will slow things down

Security has to be implemented
Can you give user A only read access and user B read\write access. What if you don't want certain users to see the salary column?

Bulk uploading of data has to be implemented
The ability to upload a file and specify the target object to load the data into has to be implemented. We get files via ftp all the time, nobody wants to parse the files locally and insert row by row.

The ability to do merge replication with disconnected handheld devices
This is not so much an issue anymore since almost all devise these days are online but if you have some legacy junk this would be nice.

The ability to restore to a point in time
You uploaded some bad or old data and hosed everything. On a SQL box you would just do a point in time restore (if you had backups that were valid of course and the correct recovery model) and you would be done.

Alerts and Notifications
You want to know every time some product level falls below a certain threshold, someone added or modified a table etc etc.

Auditing, SOX, HIPAA, GAAP and all your other favorite acronyms
Unless this stuff is implemented no public company will move to the cloud.

Bandwidth cost has to be low
There are tons of FoxPro, Access Applications Excel sheets and in house developed applications that connect directly to SQL Server, get a ton of data and then do something with that. When this is all in your office you don't have bandwidth cost. If your data lives in the cloud this could add up.
I will give you another example, we all compress our content before it is pushed to the client from our websites right(rhetorical don't answer)? Guess what? Amazon doesn't do compression because they get paid for bandwidth, compressing would lower their revenue.

You need to have the ability to profile
There has to a way to profile and performance tune your cloud database, after all that is the most fun part of a database developer, you can’t take away instant gratification


What will go in the cloud first?
The first thing that will ascend into the clouds are your typical internet applications, this will also include RIA (or as I call them Chubby Clients) that are built on Adobe AIR, JavaFX and Silverlight. Ever heard of TweetDeck? It is a twitter client written in Adobe AIR, stuff like that is perfect because you don't care where the data is stored; you just want to call a RESTful API and get your data. Websites that don’t store (or don’t care) about sensitive information are going to store their stuff in the cloud, now we have real C2C (Cloud to Cloud) since websites live in the cloud to begin with.


What will not go in the cloud any time soon?
Any app that is directly connected to a database server and does some crazy manipulation/calculation of data, the bandwidth cost might just be too high
Departmental (Small business edition) database servers or SQL Server Express edition, this stuff is cheap enough that I can’t imagine there being an incentive to move to the cloud. Shops like this usually have one person to manage all the hardware to begin with so they would not save anything.
I am also having a hard time seeing those lovely Excel sheets that are connected to SSAS cubes connecting to the cloud instead.

 
 
We cannot really predict anything unless we have some pricing. I aslo think virtualization of databases will be mainstream sooner than the cloud, one huge box 20 SQL Servers on it...only one box to manage
 
If you want you can vote (you need to register), right now it is mostly no cloud 
 
 
So what is your opinion on this, The Cloud is it vaporware or inevitable?

Published Wednesday, February 25, 2009 11:47 AM by Denis Gobo

Comment Notification

If you would like to receive an email when updates are made to this post, please register here

Subscribe to this post's comments using RSS

Comments

 

jamiet said:

"The ability to upload a file and specify the target object to load the data into has to be implemented. We get files via ftp all the time, nobody wants to parse the files locally and insert row by row."

Denis,

I wouldn't worry too much about this - I'm sure there'll be tools available. In some cases there already is - you can already download a free SSIS destination adapter for SDS - i.e. A component that uploads the contents of a SSIS dataflow into SDS.

-Jamie

February 25, 2009 12:01 PM
 

Morris Lewis said:

The medical center I work for is private, and you don't even have to limit statement about the acronym soup requirements to public companies. There is simply no way health care organizations are going to trust their data and their ability to provide their services to a third party they don't control. The risk of impacting patient safety and the loss of money and goodwill is far greater than any benefit the cloud may offer. I can see companies creating their own clouds, especially if it improves uptime or offers better disaster recovery solutions, but that's the only way I can see clouds being more than Microsoft's current "golly, look at what our research dept cooked up this week!"

February 25, 2009 12:07 PM
 

James Luetkehoelter said:

I think Morris' comment is dead on - the biggest issue with a the Cloud concept when it comes to information (not processing) is one of trust. A "cloud" implies something unknown or undefinable happening. That's why you have a network diagram that ultimately leads to a "cloud" icon if you're hitting an external network system (which isn't necessarily the Internet).

I know of many private companies that have 200MB databases that would never trust an outside agency to host it without being able to ultimately "point" to that data ("it's on server x at our location B that also has data for 1000 other customers).

Cloud computing I think is an inevitability - cloud data an oxymoron. Data I believe must reside on a discrete island, with a discrete "owner".

February 25, 2009 6:10 PM
 

Kasper de Jonge said:

Well my guess is that cloud computing will take flight soon, think MS Azure and SSDS, it will make hosting .NET websites in the cloud with data from webservices rather simple (as far as i can see)

February 26, 2009 2:39 AM
 

Jason said:

This makes this cloud more interesting to me.

http://www.theregister.co.uk/2009/02/23/sql_server_azure/

February 27, 2009 9:27 AM
 

Bikerdad said:

Perhaps its time to re-read William Gibson's <i>Neuromancer</i>.  It's not that much of a step from the Internet to Clouds to full blown Cyberspace.  What is pretty much a given in terms of tech advances is this:

If you can see it on the horizon, and it looks like a 'slam dunk', its gonna take a LONG time.  Anybody got their flying car yet?  It's the things that come out of nowhere that move fast.

"scattered clouds" is the way of the near (5-20 yr) future.  Physical security issues, bandwidth, and legal are the primary challenges.  Outside of the theoretical sciences, too much of the data that intensive crunching applications use is subject to privacy concerns, whether legal (i.e. medical data), or proprietary (does Company X really want its engineering prototype info in the cloud?).  High volume low criticality stuff will make it to the cloud first, which means entertainment.  Porn clouds, music clouds, etc.  The problem with these is the bandwidth issues.

March 2, 2009 12:33 PM

Leave a Comment

(required) 
(required) 
Submit

About Denis Gobo

I was born in Croatia in 1970, when I was one I moved to Amsterdam (and yes Ajax is THE team in Holland) and finally in 1993 I came to the US. I have lived in New York City for a bunch of years and currently live in Princeton, New Jersey with my wife and 3 kids. I work for Dow Jones as a Database architect in the indexes department, one drawback: since our data goes back all the way to May 1896 I cannot use smalldates ;-( I have been working with SQL server since version 6.5 and compared to all the other bloggers here I am a n00b. Some of you might know me from http://sqlservercode.blogspot.com/ or even from some of the newsgroups where I go by the name Denis the SQL Menace If you are a Tek-Tips user then you might know me by the name SQLDenis, I am one of the guys answering SQL Questions in the SQL Programming forum.

This Blog

Syndication

Powered by Community Server (Commercial Edition), by Telligent Systems
  Privacy Statement