Experiments, Lessons Learned, Best Practices & Tutorials
Amazon is the undoubted leader in the public cloud space today. They offer more features than any other provider today have reached a level maturity that is making them the top choice for Enterprise and the government. Their competitors are looking for ways to differentiate from them. IBM for example is touting their Enterprise experience. Rackspace is using their customer intimacy. Microsoft is using their muscle giving away tens of thousands in free credits for trying Azure. And then we have Google, who are using their performance, usability, and lastly price. Google Computer engine is a great bang for the money. Google has some very interesting features that people had desired Amazon to add for a while, and higher network performance too. At a lower price, Google does become a very compelling option. Undoubtedly, we are not the only ones who find Google a viable alternative — so does Amazon Web Services themselves. They showed this by responding to Google’s price cuts at the AWS Summit last Wednesday.
The price cuts were drastic, e.g., a 51% reduction in S3 costs, but not unexpected. After Google’s announcements, everyone was expecting AWS to respond. There were several tweets I saw including a text message I received from a friend (CEO of a stealth mode startup): “Let’s see if there is a price cut announcement.” This change was clearly welcomed by the AWS user community. Twitter was flooded with tweets. The following is the slide announcing the price cuts. You can see the newer pricing by going to the Amazon website.
What’s interesting to note is that prices of the M3 and C3 instances were cut by 30%. We have already blogged about c3 instances being the best bang for the buck today in AWS (assuming they fit your minimum requirements).
[Tweet “This is what #AWS Price cut would mean to your enterprise”]
What will price cuts mean to those who are still using on-premise infrastructure?
If you are one of the SMBs and Enterprise who were on the fence about AWS, you need to reassess. However, your new assessment ought to be done differently. When re-evaluating, keep the following in mind:
A. The costs are lower
B. Consider the bigger message here that AWS has promised that they will continue to reduce prices over time. Thus, your analysis must take anticipated savings into account. Even if today AWS is slightly more expensive than your on-premise infrastructure, it will likely become cheaper over time.
C. Heavier price cuts on the latest c3 and m3 instances imply that it is likely you will also get higher performing hardware at lower prices. Your analysis must take this into account this performance difference between your own equipment and what AWS is offering.
If you are looking for a bias-free assessment, we can help. Contact us here for a free 30-min consultation.
What will price cuts mean to Enterprise and mid-size businesses already in AWS?
Your monthly bill will reduce. However, the price cuts were not homogeneous across the board. What this implies is that if you thought m1.large was your best bang for the buck until today, you must re-assess. Contact us if you like to assess what you changes in your AWS infrastructure can allow you to take maximal advantage of the new price cuts.
What will the price cuts mean to a startup?
Your AWS bill will be lower, hence your burn rate. Lower burn rate is great news for every startup. It allows them to sustain longer thereby increasing success. This also implies that as a startup, you do not need to worry about “what will happen to the AWS costs as my company scales?” If the prices will be cut over time, you are best off just staying in the cloud for as long as possible.
[Tweet “What will the #AWS price cut mean to a #SaaS startup?”]
Four Easy Tips to get maximal benefit from the latest price cuts
1. We recommend switching all m1.medium instances to c3.large instances and m1.large to c3.xlarge. c3 instances are a clear winner in terms of performance as they have newer processors and SSDs. c3 instances are slightly higher priced but for most people switching meant an increase in their spend. Not anymore. If you switch after today, you will be able to see a reduction in cost and an improvement in performance at the same time.
2. Use S3 more than ever before. Ideas to use S3.
A. Host your static files in S3. It reduces cost and improves performance and availability.
B. Replace your FTP and file servers with S3. For example, one of our Fortune 500 clients use to push firmware updates to their hardware in the field using FTP servers: devices would use FTP to get the latest firmware every few hours. Not any more. We switched them to using S3 and HTTP, and now they no longer have to pay for the server or the resources to maintain it. Plus S3 offers versioning.
C. Use S3 for Backup/archival of data, even on-premise data like nightly backups and images. For example, one of Flux7 team members has backup all his pictures in S3 for under $1/month.
3. Switch your t1.micro instances over to m1.small. The price difference between the two has just reduced but an m1.small is more reliable and persistent in terms of performance. It is far more “production-ready.”
4. Lastly, again consider enabling self-service IT. With lower computing costs, the risk of getting a sticker shock is reduced. Yet another reason to enable self-service IT. With a reduced computing cost bill, make your Devs more productive by allowing them to use more computing resources and improve their productivity.
Question to ponder: After the new price cuts, one hour of a typical US developer costs more than running a typical AWS instance 24/7 for one month. Is it ever a wise-trade off to make a developer wait for even 1 hour for an IT ticket to be serviced?