One of the concerns often voiced about cloud computing/SaaS is the reliability of the service. What is the performance and up-time guarantee. Will the users be unable to access the application or service right when they need it most?
Over the last 10 years, there have certainly been growing pains. salesforce.com, widely recognized as the SaaS pioneer broke new ground in 2006 when it introduced ultimate transparency by vowing to post information on its website about the status of its service. The URL trust.salesforce.com provides details on the status of their services, including near real time reporting.
Another cloud computing stalwart Google now offers SLAs (Service Level Agreements) for paid Google Apps users, which guarantees service credits based on uptime levels not achieved. 99.9% being the limit in which service is acceptable.
| Acceptable Uptime (%) | Downtime Per day | Downtime Per month | Downtime Per year |
|---|---|---|---|
| 95 | 72.00 minutes | 36 hours | 18.26 days |
| 99 | 14.40 minutes | 7 hours | 3.65 days |
| 99.9 | 86.40 seconds | 43 minutes | 8.77 hours |
| 99.99 | 8.64 seconds | 4 minutes | 52.60 minutes |
| 99.999 | 0.86 seconds | 26 seconds | 5.26 minutes |
As you can see 99.9 is a far cry away from the infamous 99.999 (five 9’s) which would be barely noticeable to a user.
The fact is that outages are going to occur, there is even a website called the Cloud Computing Incidents Database (CCID) which collects and suggests that everyone contributes details about known outages of the leading vendors.
Keying in on availability as one of the major sticking points, even major SI vendors are getting into the act, evidenced by the recent launch of Computer Science Corporation’s (CSC) cloud computing services and their focus on The Trusted Cloud.
As cloud computing gains popularity, private clouds (dedicated clouds for a single corporation vs share public clouds) are being considered by internal IT groups of large enterprises as a means of more efficiently providing applications and computing power to their end users. Will internal IT hold themselves to the same level of transparency and trust, like public cloud vendors salesforce.com or even Google Apps? Or will they suffer the growing pains like salesforce.com did 10 years ago, until they cracked the code on how to deliver high performance, high availability cloud services that supports 60,000+ customers worldwide today.
Whatever cloud(s) you choose to run your applications on, transparency, reliability and ultimately trust is paramount, you can bet your bottom dollar on it.
P.S. There is providing the SLA or performance levels that your customers expect, and there is going straight to 0% up-time (i.e. going out of business) which is what happened to SaaS BI vendor LucidEra a few days ago. I’ll have more to say about that in the future, but in the meantime, read more about LucidEra Hysteria on Merv Adrian’s blog.
Very relevant point of reliability and transparency which will ultimately be a win-win situation for the technology and the vendor offering the same. But,I also wonder if it is too much too soon to expect from the cloud framework and infrastructure to hold on to such high SLA’s? Are the recent outages of Twitter and Google anything to go by? How secure are these models is the other question along with the relevant ones pointed above.Cloud Security cometh.
Though the above table gives the statistics w.r.t the overall cloud computing, I wonder how the distribution is within the HPPIE – Hybrid, Private, Public, Internal External(or is it HPP?)- models of the cloud? This would lead to a best of the breed architecture for this framework and help build the TRUST.