Get started on Windows Azure: Attend “Meet Windows Azure” event Online

Standard

On June 7th 2012 – there’s an online event called “Meet Windows Azure” where Scott Gu and his Windows Azure team would introduce the Windows Azure platform. You can register here: http://register.meetwindowsazure.com/

If you’re planning to attend – there’s a very interesting tweet-up planned called “Social meet up on Twitter for MEET Windows Azure on June 7th” – All you have to do is follow #MeetAzure, #WindowsAzure on Twitter & Interact! Simple!

There’s an unofficial blog relay, if you write a post – Tweet it to @noopman – Here is the Blog Relay:

Played with Microsoft research “Project Daytona” – MapReduce on Windows Azure

Standard

Recently, I played with Project Daytona which is a MapReduce on Windows Azure.

It seems like a great “Data Analytic’s as a service”. I tried the k-means and the word-count sample application that comes bundled with the project run-time download: http://parasdoshi.visibli.com/share/z14Ty2

The documentation along with the project guides you in a step by step fashion on how to go about setting up the environment but for those who are curious, here is a brief description on how I setup the environment:

1) Uploaded the sample data-sets to Azure Storage

2) Edited the configuration file (ServiceConfiguration.cscfg) to point to correct Azure Storage

3) Chose the Instance size and the no. of Instances for the deployment

4) Deployed the binaries to Windows Azure (.cspkg and .cscfg)

5) Ran the Word Count Sample

6) Ran the K-means Sample

Conclusion: It was pretty amazing to run MapReduce on Windows Azure. If you are into BigData, MapReduce, Data Analytic’s – then check out “Project Daytona”

That’s about for this post. And what do you think about Project Daytona – MapReduce on Windows Azure?

One more way to run SQL Server on cloud: SQL server on AWS RDS

Standard

Up until April 2012, the only way to run SQL server on cloud was “SQL Azure”. But recently AWS announced SQL Server on Cloud. Good news? Probably. it’s always good to have more than one option. So for those who are new to world of AWS, here are few tips before you get hands-on:

1) The way RDS works is that you spin up “db instances”. So here you specify the machine size that would “power” your database. And remember that the type of instance you choose would directly affect your bill.

2) Spend some time understanding the billing structure. Since AWS gives you lot of options – their billing structure is not simple. Don’t get me wrong, I am not saying that lot of options in AWS is bad. it’s just that the billing is not simple and it’s not one-dimensional (there are various dimensions that shapes your billing structure). And why should you invest time? because in the “pay – as – you – go ” model it would directly affect your Bill.

3) understand costs like: cost to back-up database PLUS data-transfer cost.

4) Understand the difference between “Bring your OWN license” and “license included” (Express, Standard and web only. Currently enterprise edition not included here) model in RDS SQL Server

5) and unlike SQL Azure, RDS SQL Server charges on a “per hour” basis.

Note the date of this post: 15th may 2012. Things change very fast, so readers-from-the-future please refer to official documents.

BTW, here are the few blog posts from the web-o-sphere:

1. Expanding the Cloud for Windows Developers

2. First Look – SQL Server on Amazon Web Services RDS

3. Official resource: AWS RDS SQL Server

That’s about it for this post.

How do you reduce the network “latency” between application and SQL Azure?

Standard

I was at SQL Rally recently (10-11 may 2012) where I happened to have a nice talk about SQL Azure with a fellow attendee. They were considering porting their database (that supports one of their apps) to Microsoft’s cloud service. One of the concern they had was “How to reduce the network latency between SQL Azure and their App?” And Since I knew the solution, I shared it with that person. I am sharing it here so others can benefit too.

Now one of the first question that I asked the attendee was: Are you also porting your app along with the database to Azure?

Turns out, They were considering to host the app on Azure cloud too. So technically that’s called a “Code Near” scenario – And in this case, the application and the database both *should* reside in the same data-center. if you do so, the network latency between your app and the database is minimal.

Now, if you have your app on-premise and you are considering SQL Azure, then select the data-center location that has the minimal network latency between your app and SQL azure. Technically it’s called Code-Far scenario I have written about one of the ways you can do so, here’s the URL: Testing latency between client and SQL Azure via client statistics in SSMS

That’s about it for this post.

Official Resource: SQL Azure and Data access

Business operation challenges faced by SaaS provider & cloud provider

Standard

I just researched few business operation challenges that are faced by SaaS (software as a service) provider & cloud provider, I am sharing what I found here:

  1. SaaS and cloud computing companies have wafer thin margins (a.k.a lower margin businesses). According to SaaS update (April 2011 2008, pg 15), Their operating margins are just about 3%. You can do quick search to know past operating margins for the SaaS and cloud provider of your choice.
  2. SaaS and cloud computing companies have to manage the customer churn rate which means these companies have to put efforts in retaining customers. Churn rate is a measure of the number of individuals moving into or out of a collective over a specific period of time. Now managing churn rate is a challenge because retaining customers is difficult and time-consuming. And customer retention is important because in the pay-as-you-go nature of subscription businesses, the customers pay only if they continue to use the service. Thus managing customer churn rate is a challenge for SaaS provider and cloud provider.
  3. Cloud provider and SaaS provider needs an upfront investment to build sales capacity needed to grow the business over time. And so if they invest less, they do not have enough sales capacity and hence they will miss growth opportunities. Thus they have to make a careful trade-off between fast growth and high cash burn rate.

What’s the widely accepted definition of cloud computing?

Standard

Cloud computing is no more a “wild west”. Their seems to be a consensus on what cloud is – And more importantly what cloud is NOT. So what’s the widely accepted definition of cloud computing? Well, It’s the NIST definition of cloud computing.

It has three pieces to it, Here they are:

  • Characteristics of a CLOUD
  • Delivery (/Consumption/Service) Model of CLOUD
  • Deployment Model for cloud

Please search for “NIST Definition of Cloud Computing” and you’ll get to a (latest version/draft) short PDF that is worth reading – takes just 10 odd minutes but super helpful!

For your convenience (and Mine) – Here’s a visualization of the “data” that’s in the NIST’s definition document:

NIST definition of CLOUD

NIST's Widely accepted definition of CLOUD COMPUTING

If you want fun way to remember the definition, please go to: Cloud Computing is AWESOME (OSSM) :: Defining Cloud Computing the Urban Dictionary style! (It’s goal is to help you remember the characteristics of cloud, and i make fun of NIST’s definition – not because it’s not great but it seems academic which may not be the best way to explain cloud to non-technical or business person. So a fun way to remember cloud helps you have a sophisticated discussion with someone who is not completely familiar with cloud. Hope that blog posts is of help too)

And That’s about it for this post. Your feedback is welcome!

And Let’s connect! I Look forward to Interacting with you on any of these people networks:

paras doshi blog on facebookparas doshi twitterparas doshi google plusparas doshi linkedin

A nice slidedeck on SQL Azure reporting: #SQLPASS Business Intelligence Virtual Chapter Slides and References by Simran Jindal

Standard

recently I attended a nice SQL PASS Business Intelligence Virtual chapter’s presentation on “SQL Azure reporting” by Simran Jindal. For anyone interested in the slide-deck, re-blogging the post – please go to: http://simranjindal.wordpress.com/2012/03/08/sqlpass-business-intelligence-virtual-chapter-slides-and-references/ ; Also the post has few great resources, check them out if you are interested in learning about SQL Azure reporting.

I have written level 100 articles before and here are the links if you want to browse through the basics first:

What is SQL Azure reporting?

Getting started with SQL Azure reporting

Gartner’s Hype cycle for CLOUD COMPUTING: 2008 – 2009 – 2010 – 2011

Standard

In second half of 2008, I had taken a class on Distributed systems. As a part of that class, We were asked to write a paper on the subject of our choice. I had chosen  the subject that focused on answering: How is cloud different from traditional approaches? My aim was to figure out why was CLOUD getting all the “attention”. And i chose this topic because, at that time [2008], I used to follow few blogs (including Gartner) at that time and almost all of them seemed to be “heavily promoting” cloud. So I asked myself “why is world going crazy about cloud?” “What makes cloud different from traditional approaches?” And I decided to find answers to this question and convert it into a term paper for the class. And during the term paper assignment, I got a traditional Hosting account and deployed one of my project there. And compared it with what literature on cloud computing had to say at that time. This approach was super helpful because I had a benchmark on what traditional hosting felt like and then what cloud computing enthusiast had to say about why is cloud better than traditional approaches.And after getting my head around that concept, it’s been fun following how “cloud” has evolved over time. One of the best ways to recap it is to analyze Garnter Hype cycle. To this end, Here’s a recap of where Gartner Places Cloud Computing (and its different flavors) on hype cycle from 2008, 2009, 2010 and 2011. Here’s a recap:

2008:

Gartner Hype Cycle for CLOUD COMPUTING 2008

2009:

Gartner Hype Cycle for CLOUD COMPUTING 2009

2010: (Note the presence of different flavors of cloud on the hype cycle)

Gartner Hype Cycle for CLOUD COMPUTING 2010

2011:

Gartner Hype Cycle for CLOUD COMPUTING 2011

Image Courtesy: http://www.gartner.com

And That’s about it for this post. Your feedback is welcome!

And Let’s connect! I Look forward to Interacting with you on any of these people networks:

paras doshi blog on facebookparas doshi twitterparas doshi google plusparas doshi linkedin

Great to see neck to neck price competition between Azure and AWS

Standard

[I am referring to price reduction announcements by Azure and AWS in Early March 2012]

This week, Microsoft announced price reduction for Windows Azure compute and storage. Read more here – And even AWS announced price reduction too. Read more here. So it’s great to see neck to neck competition between Azure and AWS. And this is what is great about an economy that is not “monopolized”. To sustain, both Azure and AWS need to compete PLUS they compete with other cloud vendors that are out there too. Hopefully, other cloud vendors would also come up with an aggressive pricing strategy. Also, such competition among vendors would spur innovation and we would see “more features” coming out their factories more often than before (Yay!). And guess what, Who’s the winner? WE the CUSTOMERS!

Also, I noticed a pattern in both announcements that said “We are glad to pass along the savings to customer”. This seems to pointing to the fact that since cloud computing adoption has increased and cloud vendors benefit from Economies at scale – They pass the savings to the customers. And this acts as a catalyst for more adoption!

This seems like an awesome circle:

cycle of cloud vendor price reduction and adoption

I understand that CLOUD ADOPTION is just not triggered by “price reduction”. That’s not what i mean here. But price reduction can sure act as catalyst. Hopefully, in not so distant future, the cloud prices would be so lucrative that setting up private data-centers would be a “thing of past” (unless you are governed by laws to have your own data-center or some other policy that restricts you from cloud adoption).

And I am also not comparing the price reduction “directly” because it would be like comparing Apples and Oranges. But what I hoped to point out was that we as customers would see more price reduction, more features, better experience because of the neck to neck competition among cloud vendors.

And That’s about it for this post. Your feedback is welcome!

And Let’s connect! I Look forward to Interacting with you on any of these people networks:

paras doshi blog on facebookparas doshi twitterparas doshi google plusparas doshi linkedin

[video] The day it all began for SQL Azure

Standard

SlideDeck: Introducing SQL Server Data Services

 

Video: Introducing SQL server Data services:SQL Server Data services SSDSHere’s the Announcement: Announcing SQL Server Data Services

It’s interesting to me how SQL Azure has been evolving at an amazing pace! And we forget that this service is relatively new – After all it was announced in April ’08 at Mix. So it’s good to see that SQL Azure (previously known as SQL server Data services) is growing rapidly. It seems like it’s a start-up – no kidding. And that’s great!

Check out above video – if you like digging through “historical” material.

That’s about it for this post.