AWS Summit 2011 – ‘my other computer is an EC2 instance!’

Same famous entrance shared by Werner Vogels and Tony Blair. AWS Summit and the Iraq Inquiry were both held at Queen Elizabeth II Conference Centre, London

Smart421 attended the AWS Summit 2011 on Tues 14 June at the Queen Elizabeth II Conference Centre in London and I wanted to share some ’soundbites’ that caught my attention.

I guess there are a succession of these that are deliberately trotted out at events like this but all the same they are worth repeating, and I’m sure I will trot these out myself in the coming months .

Hopefully I will have time to share some of the other information that I gleaned from the ‘Base Camp’ track that I followed during the event, and the 7 key themes that Werner Vogels (Amazon.com CTO – and yes he was there!) sees driving the Cloud, on a subsequent blog post.

 

Anyway here are a few beauties!

• From Nov 2010 all Amazon.com web traffic is now being served from EC2 instances – talk about ‘eat you own dog food!’
• A large media company has 1500 VM’s of which 5% are currently in the Cloud and the rest on-premise – by 2013 the figure will be 50% in the Cloud – wow big transition, and yes he did mention some internal pain points!
• Same media company pushing out speculative websites that may or may not attract interest (i.e. go viral or die) – traditional on-premise time/cost model = 7 days to build & $75.40 a day to run VERSUS  Cloud (presumably AWS eek) = 3 hours to build & $27.60 a day to run =========== no brainer!
• Agility using AWS was demonstrated by a software house (don’t worry I won’t mention their name here – pah) – in the last 7 days 41 developers, 882 commits, 400 deploys for 40 projects – that’s a deployment every 5 mins apparently, and in my book that’s also bragging for braggings sake – good going guys.
• And finally AWS themselves still see all that has been done so far as DAY 1 – jeez that is scary and can’t wait for DAY 2……..

:: Stop Press ::
Check out the blogs by Robin Meehan on this and other subjects, please click here.

 

Amazon RDS for Oracle 11g Database Released!

Nearly a month behind original indications from the AWS Oracle guys but Oracle RDS is here (released 23rd May).

http://aws.amazon.com/about-aws/whats-new/2011/05/23/amazon-rds-for-oracle-database/

Below are some quick calculations for a months continuous usage (excluding storage) in EU Ireland for the On Demand model. You will see the cost doubles each time you in theory double the power

On Demand (License Included)	High Mem/Extra Large DB Instance	$699.36 per month
	High Mem/Double Extra Large DB Instance	$1,398.72 per month
	High Mem/Quadruple Extra Large DB instance	$2,797.44 per month
On Demand (BYOL)	High Mem/Extra Large DB Instance	$550.56 per month
	High Mem/Double Extra Large DB Instance	$1,101.12 per month
	High Mem/Quadruple Extra Large DB instance	$2,202.24 per month

The Reserved model still remains significantly cheaper over a year period – $305.05 per month as opposed to $699.36 but with upfront year 1 cost of $1,850 – so year 1 costs £5510.48 for reserved as opposed to $8392.32 for on demand for the example of high mem/extra large instance.

Some of our Customers have been keen to see progress on the AWS-Oracle roadmap and whilst any progress will always be welcome news, one of the key observations on ‘first-glance’ of the Oracle RDS service is that there is a multi-AZ option but that the only current available selection is ‘No’ - this is not the case for the MySQL RDS service. This will be a limiting factor particularly for our enterprise Customers when considering a highly resilient/highly available database solution. Some digging is required into this but it is likely to be a consequence of the underlying AWS architecture and the ‘new ways’ of thinking in terms of having a ’shared nothing’ architecture - this currently prevents mutli-node Oracle RAC implementations on AWS.

Based on the recent AWS outage  (see previous blog posting here) and the re-enforced message to implement your solution across AZ’s to ensure high resilience/availability, then this Oracle release may not be enough to impress enterprise-level Customers – it may be a case of  continuing to ‘watch this space’ on the progress along the AWS-Oracle roadmap……

 

Why does AWS stress security in the Cloud?

Richard Holland, Operations and Delivery Director of Eagle Genomics addresses delegates at the 2011 AWS Tech Summit, London 17 March.

The AWS Technical Summit in London 17 March [see http://aws.amazon.com/aws-tech-summit-london-2011/] was very worthwhile and no one could fail to notice that AWS themselves were taken by surprise in the exceptional attendance levels. Along with colleagues, Smart421 attended as AWS Solution Providers.

One of the key takeaways for me was how AWS continues to be highly responsive to the market. As well as bringing features to market in rapid succession, they have also listened and replied convincingly on the recurrent obstacle of security in the Cloud.

In fact, in addition to AWS’s already comprehensive security stance [see http://aws.amazon.com/security/] it emerged that one of AWS’s customers, Eagle Genomics based in Cambridge UK, had also permitted two independent IT consulting firms, AT&T and Cognizant, to perform ethical hacking on their AWS instances (permissions obviously required in advance).  The outcome?  Both firms reported that it couldn’t be done.

For hardcore sceptics (aren’t we all at heart), AWS say they have a list of partners that provide services around ethical hacking. This means that if your business case warrants a belt and braces approach on security, it’s possible to engage IT consulting firms these kinds of trials dedicated to your instance or instances in the Amazon cloud.

In reality, with some notable exceptions where systems are understandably internalised, both physical security and digital security offered in AWS is far in excess what the majority of organisations are able to provide for themselves. I’m certainly not alone in thinking that a far bigger risk for enterprises resides in the portability of data (e.g. copies held on company laptops, CDs, USB keys, etc) than in hacking instances of Cloud computing, particularly those on AWS.

It would seem that the objection around security in the cloud is being steadily eroded away. About time.

 

Amazon Web Services in Japan – but I don’t get the pricing

AWS continued their expansion the other day by announcing a new Japan region, hosted in Tokyo.

What I don’t quite understand is some of the pricing differences. I can understand that bandwidth might be different in different territories, and maybe the price of hardware (local taxes maybe? different shipping and local labour costs etc?), but if you compare the EC2 EU region with the new APAC-Tokyo region, you can see that whilst the Windows costs are the same, the Linux costs are higher in Tokyo.

As there should be no software license cost for the Linux instances, this seems a bit weird. All I can think of is that the Microsoft SPLA (Services Provider Licenses Agreement) that AWS have managed to negotiate with Microsoft happens to be cheaper than the EU region and exactly offsets the other higher costs.

 

Amazon Web Services inter-AZ latency measurements

Continuing our fascinating series (sic) of Amazon Web Services related latency measurements…we’ve already looked at the round trip time between the UK and the US vs the EU region, so now our attention has turned to the latency between availability zones (AZs) in the EU region.

The network latency between AZs is critical to designing and implementing fault tolerant applications on AWS, as the design assumption is that synchronous transactional data replication is always feasible, and you can seamless fail over from one Relational Database Node (RDS) node to another “standby” replica in another AZ etc. So we thought we’d measure it!

One of my colleagues collected some measurements between 4 x Linux t1.micro instances (2 running apache & 2 running http-ping scripts) between EU zones. The scripts were scheduled to run every 5 mins (via cron) with each executing 20 x http-ping requests and returning the average response time in milliseconds. We also measured the latency of the http-ping requests via both the private and public addresses of the corresponding web server.

As a control measure, we also measured the average latency for “localhost” to respond, to allow us to eliminate the web server response time from the measurements. This worked out to be as follows:

• Roundtrip to localhost – round-trip min/avg/max = 0.9/1.0/1.3 ms
• Roundtrip to a public IP in the same AZ – round-trip min/avg/max = 1.3/1.4/2.7 ms

And here are the results of measurements across the AZs – first for the private IP addresses (click the image to view in full size):

…and for the public IP addresses:

So in summary, the roundtrip between AZs using public IP addresses works out to be about 4ms minimum, and once you take off the minimum 1.3ms experienced between public IPs in the same AZ, and dividing by 2 (as it’s a roundtrip), then the latency between AZs in the EU region works out to be about 1.35ms minimum. Pretty quick really…

 

AWS London User Group meeting

I attended this Amazon Web Services user group meeting for the first time this evening as I was also in London today for the first day of the Cloud Expo Europe conference. When I say it was a meeting…it did take place in a pub

It was great to meet some like-minded people and kick around a good collection of subjects. Strangely no women came over to the table to see what this bunch of nerds were talking about. That’s probably because the topics included:

• Why you might use AWS over Rackspace
• IPv6 support
• AWS instance failures
• AWS vs Microsoft Azure
• When AWS is a competitive platform for bog standard web hosting even if the web site load is not peaky at all

I won’t tell you what the outcomes of these discussions were….you’ll have to ask me or attend the next one to find out! One outcome I can report is that that some beer and beefburgers were consumed.

As AWS adoption inevitably grows in the UK then this group can only get stronger, so I’ll make sure I try and attend the next event. The man who made it all happen was Craig Box, so top marks to him for organising it. He also handed out some copies of Jeff Barr ‘s book  “Host Your Web Site In The Cloud”, so if any Smarties want to have a read, just let me know.

 

In the end there is only (well…mainly) PaaS

Amazon Web Service’s announcement yesterday about their AWS Elastic Bean stalk offering was a significant step in their inexorable movement up the stack from IaaS towards PaaS (platform as a service), offering a Java PaaS based upon a Apache/Tomcat/Linux stack.

At the same time, the SaaS players are moving in the opposite direction, as demonstrated by Sales force.com’s recent acquisition of Heroku, the ruby PaaS provider.

Up until now, I’d have argued that PaaS has been the least mature of all the “aaS” siblings, with significant risk of lock-in, vendor failure risk etc, but this is rapidly changing and PaaS will become the main battleground between all the vendors as the SaaS players make their offerings more and more configurable/flexible and the IaaS vendors try and simplify their typically quite techy offerings for a wider audience.

There’s good further material of this topic on Krishnan Subramanian’s blog.

 

Which Amazon Web Services region should you use for your service?

Imagine that you want to deploy some components entirely within one Amazon Web Services (AWS) region – which region should you choose? Well, US-East is cheaper than EU-West (e.g. by 11.8% for the smaller on-demand instances), so that’s quite attractive – if your deployment is not tied for legal reasons to a European deployment. Obviously for most UK-centric organisations, the US is further away (!) and so network latencies should be greater…but are they? It seems obvious but before we ruled out a potential 10%+ saving for our customers, we thought we’d do some measurements.

And the answer is…

us-east-1d, average round trip time = 231ms

eu-west-1a, average round trip time = 96ms

This was based upon using http-ping to check the average round trip time of HTTP requests from our Ipswich development centre in the UK. So phew, the laws of physics still hold true even in a cloud world

 

Amazon Web Services SimpleDB – ready for primetime?

AWS‘s SimpleDB offering offers basic name-value pair data storage, i.e. it’s not a rich as a relational database. AWS describe it best, so I’ll just repeat what they say – “Amazon SimpleDB is a highly available, scalable, and flexible non-relational data store that offloads the work of database administration. Developers simply store and query data items via web services requests, and Amazon SimpleDB does the rest.“

So, as an AWS Partner, would we recommend its use?

Well, first I should say that SimpleDB is still officially a beta, so it’s a bit unfair to take pot shots at it – AWS have a great track record in bringing new features to market and then rapidly enhancing them – have a look at their blog and you’ll see what I mean, the rate of change is really quite scary. Note that there is no support for SimpleDB in the Amazon Management Console (not that you need very much).

SimpleDB is great for certain use cases, but has a significant weakness that there is no simple way of backing up a SimpleDB domain – you have to use the APIs to pull all the data out from a domain (aka table) and stick it somewhere, e.g. in S3. You’ve also got to write your own “recovery” mechanism. AWS manage resilience for you by maintaining multiple copies of the data etc, but obviously that is NOT the same as a backup – it won’t protect against finger-trouble for example.

This is a well known functionality gap, and various people have come up with commercial solutions to fill that gap, e.g. BackupSDB. So whilst I really like the idea of SimpleDB for certain (many!) data storage/retrieval use cases, I think I’d tend towards using their Amazon RDS offering even for simple data structures as then I have an out of the box backup approach – unless I needed absolutely ludicrous read performance/scalability or the cost advantages justified dealing with the backup issue hassle. A knowledgeable Smartie colleague of mine pointed out that for one of his AWS use cases he can recreate all the data in his SimpleDB domains from the original data source, so I think in this scenario SimpleDB is a good fit even in its current beta state.

To find out more, there’s a good discussion of the pros and cons of SimpleDB vs RDS here under the heading “Choosing an AWS Database Solution”.

 

Oracle Applications supported on Amazon EC2

Jeff Barr has posted some info here on the AWS blog which provides links to other AWS pages describing how some core Oracle Applications products are now certified and supported on Amazon EC2, including eBusiness Suite, PeopleSoft, Siebel and Fusion Middleware. It feels like Oracle have joined the party at last, and not before time. The key barriers to sticking Oracle Applications on AWS before were that:

• You had to create/customise/maintain your own AMIs (a hassle but doable)
• More importantly, when you ring Oracle with a support issue, they might not play ball

The removal of the second barrier is the key thing. Some barriers still exist – customers still have to be happy with the data security aspects of public IaaS hosting, so that’s still a barrier, maybe mainly an emotional rather than real one. Also, Oracle Real Application Clusters (RAC) deployments are not currently supported – so that’ll put most enterprises off for now I’d have thought though.

It’s been almost exactly a year since Larry Ellison’s infamous rant against cloud computing. And clearly a year is a long time in the world of cloud computing…