From 1 to 100 Vm's & beyond

So you started playing around with amazon service, deployed & used the AWS Free Tier. Slowly you & team became familiar with the API, consoles, building a custom AMI, understood zones/regions & were totally amazed by the coverage on twitter & blogs. Most importantly you paid in cents!

It took hardly a few days for you to get hold of the nuts & bolts involved. Your manager is happy to see an AWS expert on the team & your role transparently changed to an Infrastructure Architect. Business units started requesting to put marketing web sites, surveys & more. Sales started requesting for more demo instances & trial stack to hand over to prospects for a faster sales conversion.

Spinning up a web stack was never an issue. Add the flexibility of EBS, RDS, S3  & handling a new app deployment  was an easy breeze.

Dev teams are happy to have a cooler config to play with on-demand.
QA & testing fellas were never so happy with black/white & other vibrant testing schemes.
Business teams are happy with IT teams for the first time in history.
Sales guys are closing on sales like never before.

You a.k.a "Infrastructure Architect" are talk of the office & even dated that girl from the corner desk.

You must have noticed but seems like you got plenty of VM's & other virtual elements on hand already. You started with a 1 & phew! its grown to a handful. In some cases you are handling 100's of VM's, you now have to keep track of growing EBS volumes & snapshots to be taken & more.

Your time is being eaten by mundane tasks to keep your CLOUD up & running not leaving enough time to look at granular security & optimization issues. Nevertheless it seems like you need a tool which can take this burden off your shoulder. 

Tool which can

1. Track & monitor your AWS consumption
2. Can recommend optimization & cost saving strategy
3. Can alert when things hit thresholds
4. Can keep you informed on overall billing across accounts
5. & more....

Check out Attribo Cloud Insights, it will change the way you look at Amazon AWS, give more bang to your buck & of course save you time & money.

Attribo Cloud Insights is currently inviting you to join our early adopter program, for more details & sign up visit here www.attribo.com/signup.html


- Vinod

My cloud, your cloud, our cloud

Time has arrived for companies to look at cloud strategy as a collective effort. Public cloud has been looked upon as a means for developers & small number of evangelists within a company to convince the benefits & on -demand channel. Within your IT there are teams those are looking at cloud from their own perspective & channelizing the benefits to higher ups.

Come 2012 & beyond this is already changing for all & going forward it will become a strategy for Business IT to align their interests with IT all together. Roles are already getting formalized those will be responsible for the cloud strategy within IT & the CIO would have a direct role to put a plan together with. DevOps, Cloud Infrastructure Architect & others roles are already been considered for the NEW ON-DEMAND IT.

So great news that we are moving towards "Our Cloud" paradigm & it will bring in new challenges on the way.


- Vinod

Cloud goes mainstream

2006 was when I got hooked on to this new age concept of on-demand consumption. Things are not the same anymore & has helped one to innovate at a rapid pace & aggression. Successful start-up stories on how one achieved a hockey stick growth & repeat customers with the SAAS offering built on top of public infra (IAAS)  has helped foster growth.
Remember how we used to go through a week long approval cycle to just commission a demo box.
6 yrs have already passed & finally cloud is taking strong roots at both start-up as well as enterprise levels.

Kudos to Amazon for putting things together in a never before manner  & are a trend setter in true sense. Even a high school kid can innovate & build an iPhone app & scale with mom-pop pocket money.

Still business challenges are there but all have a better chance to crack the impossible & think of being the next Google or Facebook or Zynga's of the world. In fact Facebook's growth & coverage is so rapid that the IPO might actually hurt them as there are limitations as to how far can you grow, unless they launch a rocket & reach out for an intergalactic mission of spreading Facebook universally.

- Vinod 

DevOps more popular in start ups, Attribo DevOps Survey Results #1

This is the first of several postings on DevOps survey results.  The survey consisted of a detailed study of 51 DevOps job postings taken from multiple US job sites, analyzing 30+ parameters on categories such as company information, responsibilities, skills needed and tools used. 
The consideration behind this approach was that a job posting and description provided quite an authoritative statement of a company's intentions on DevOps, i.e. they are literally putting their money on it.

OBSERVATION #1:  Start ups are hiring DevOps engineers at almost twice the rate of established enterprise companies.  Is this a clear indication that DevOps’ is more prevalent in start up cultures? Welcome your input and anecdotes.

 OBSERVATION #2: The SF Bay Area is the most popular region for DevOps jobs.

Locations that had only one posting were included under "Other".  Clearly DevOps is a national and global phenomena.  What factors do you think are at play in the numbers above?

We will post additional observations in subsequent postings on other variables such as application type, cloud environment used, tool chain, on-call requirement etc.  Your input and observations are welcome.

About the survey:

1. This is not a scientific survey.  Sample size was 51 open job postings collected from various job sites in early April 2012.
2. 30+ parameters were analyzed and they fell into categories such as company information, responsibilities, skills needed and tools used
3. Only one DevOps job posting per company was considered
4. Some postings were discarded because it was not possible to establish key information such as company size, location etc.  Postings by job agencies were also discarded.
5. This survey reduces “Measurement Error” - biases caused due to leading questions and due to social desirability, which occurs when respondents feel pressured to answer questions a certain way in order to fit into a club, in this case the super cool DevOps club!
6. However, this survey is not without bias.  It has 'Sampling errors' since participation was limited to companies that posted a job, some 'Measurement error' is possible because the job poster may have missed listing all job requirements in the job description.

Attribo Cloud Insights : Quick stats so far

It's been a hectic few weeks keeping in sync with our early adopter prospects. Excited to see things shaping up & helping our users on a daily basis.

Thought of sharing  few key stats based on the participation so far & here it is:

1. Availability Zone : Virginia
   Virginia by far leads across. Primary reason we identified  it to be the cheapest & cost effective zone apart from it being the default choice by AWS when one commissions the resources.
   Attribo Cloud Insights offers a scorecard for you to compare regions/zones to figure out if costs can be saved by shifting to a lower cost region. California being the following choice by most & is catching up with Virginia rapidly.
   One interesting point which came across is also because of plenty of AWS dedicated early & growth stage start-ups being on the US-EAST coast. here is a run down for you to consider 

inux/UNIX / Windows

	$0.32 Virginia US East	$0.32 Oregon US West	$0.36 N.California US West	$0.36 Ireland EU	$0.36 Singapore Asia pacific	$0.368 Tokyo Asia pacific	$0.46 Sao Paulo South America
	$0.46 Virginia US East	$0.46 Oregon US West	$0.46 Ireland EU	$0.46 Singapore Asia pacific	$0.46 Tokyo Asia pacific	$0.5 N.California US West	$0.6 Sao Paulo South America

1. VM Config/Type : m1.large
   By far m1.large is being used in most production environments. Sometimes it seems to be an over provisioned choice but hey gives a little buffer to suffice for any spikes.
   c1.medium config recently added by amazon is catching up on this front & seems a reasonable choice by most. 7GB is a sweet spot for most web apps those can be put in production till one gets TC coverage or has stable spikes. But hey who cares AWS has plenty of options to scale out on-demand.
   
   

   Large

   4 EC2 Compute Units (2 virtual cores with 2 EC2 Compute Units each)
   Instance 7.5 GB of memory
   850 GB of local instance storage
   64-bit platform
   High
   m1.large
   
   Here is a cost comparison for Large instance for UNIX/Windows
   
   
   
   Linux/UNIX
   

   $0.32/hr

   $7.7/d | $53.8/w | $215.0/m | $2580.5/y

   
   

   $0.46/hr

   $11.0/d | $77.3/w | $309.1/m | $3709.4/y
   
   Windows
   

   $0.46/hr

   $11.0/d | $77.3/w | $309.1/m | $3709.4/y

   
   

   $0.6/hr

   $14.4/d | $100.8/w | $403.2/m | $4838.4/y

1. 
   
2. Architecture : 64 Bit
   Its been proposed by all & most are aware of the benefits, we still found plenty of 32 bit VM's being in use. Im sure all have a plan to move over to 64-bit. primary reason also would be to support SMART SIZING on-demand & it can be achieved with 64bit one's only

More as we progress. 

- Vinod 

For IT, Business Managers & DevOps

A common view into your Cloud IT

From pay-per-use to pay-per-abuse

Pay-per-use rocks!
I'm sure we all are already familiar with the on-demand & pay-per-use model as applied to IT. Is has been of tremendous help while executing your IT strategy & has saved you in $1000s by now. No new hardware to be commissioned or salvaged, compute & storage is being ordered as we order pizza for lunch. You select a topping, crust, flavor & boom its ready within few seconds.

This new age consumption model has given choice & power directly in hands of everyone with-in your business & depending on long request/approval cycles are monsters in the past now.

On the hind-side it has bought in a new issue in terms of  resources being used & paid on an hourly basis & is leading towards a Pay-per-abuse issue. Recently had a conversation with few key IT decision makers & the first & foremost point they wanted to discuss was to "How to curtail & control" resources abuse.

Remember you are paying by hour & if your team is not proactive enough you could end up paying good amount of your budget which could have been optimally used elsewhere. With high end compute available in seconds its easy to spring up as many virtual resources.

Sure you will get details of it at the end of month but that's too late already.

Glad to introduce Attribo Cloud Insights to you & you are welcome to see by yourself if it applies to your Cloud IT.

- Vinod