Image by ivanlanin via Flickr
I am revisiting a topic I discussed in my recap of Glue Conference Day 1.
It is a question about where do we place the responsibility when a
Cloud service goes down. With more and more enterprises jumping into
Cloud Computing bandwagon, this question becomes all the more
important. It is time for the Cloud vendors to evolve a strategy that
will make customers trust Cloud Computing.
As
more and more people are using Cloud Computing, whether it is Amazon
Web Services or Gmail or Google/Zoho Office Suite (Disclaimer: Zoho is
the sponsor of this blog but this is my personal opinion) or Evernote
or Freshbooks, users expect them to be available at all times. In fact,
the lure of the Clouds is the reliability due to the geographically
redundant infrastructure offering ubiquitous availability of data and
apps from any location and on any device. Emphasizing this fact creates
an impression in the minds of the customers and make them yearn for a
100% availability of Cloud based services. We can’t blame them. Part of
it is due to the propaganda by evangelists and marketers about the rock
solid reliability of Cloud based services, resulting in such an
impression among the Cloud users. Such an expectation puts a burden on
the Cloud providers to either offer a 100% uptime (huh!!) or offer a
trusted solution to update the users about the cause of any downtime
and the expected time of resolution.
Different
vendors try different options to solve this problem but the prevalent
feeling, especially among the potential enterprise customers, is that
it is not enough. Some smaller vendors like Sugarsync use Twitter and
blog posts to update users about the downtime and the expected time of
resolution. Companies like Salesforce.com, Google, Zoho and many others
use the dashboard approach to keep the users updated about any downtime
in their services. They also follow up such downtimes with a blog post
explaining the cause.
Even though these
solutions are better than keeping the users in dark, it is not enough.
It reminds me of the time I lived in Florida, when several hurricanes
caused havoc in the state one after the another. We were without power
for almost a month that year. But, we were not worried or upset about
it because the, then, Florida Governor was constantly updating us about
what was happening by personally talking to people through radio (and
television for those who were lucky enough to retain power). He was
updating people about the weather conditions, the impact of the
hurricane and how they are responding to it. This action gave
confidence to people that the state was doing everything needed to put
them out of misery and, also, informed them in clear cut terms that the
buck stopped with the Governor. If something went awfully wrong, they
knew who should be held accountable. Then, compare it with Hurricane
Katrina and its aftermath. I am not here to talk politics or blame
individuals. I just want to highlight the concerns of Cloud users
during downtimes. I just want to highlight what will give them
confidence during such downtimes and how they can be made to trust the
Clouds in spite of the downtimes.
It is
time for Cloud providers to come up with solutions that will keep the
users’ trust intact. It is time for Cloud providers to go beyond the
green and red dots in their dashboard and do something meaningful. It
is time for Cloud providers to realize that such green and red dots
won’t give the necessary confidence to enterprise customers and push
them into the Cloud bandwagon. Users need a throat to choke when things
go wrong and it is the responsibility of Cloud vendors to offer one.
This is a great article.
Though I am not familiar with a Google Dashboard. “Companies like Salesforce.com, Google, Zoho and many others use the dashboard approach to keep the users updated “
[..] While one may be able to define the common architectural and infrastructure layers of the cloud with common language, applying them to specific business problems will take a much more flexible construct. The early innovation in the cloud seems to be in the social networking space where the use of APIs and cloud based service delivery platforms are comfortable to engineers. But I was amazed at how cavalier those with a focus on social were about applying cloud based principles to the enterprise space. I asked several people at the conference who would be accountable for quality of service, uptime, reliability, etc. to the the enterprise customer of small business owner in a loosley coupled application environment where multiple “services” were integrated into a single application environment (for example a third-party calendar application integrated into a CRM application. The answer I typically got was; “well the calendar is a small application so its no big deal if it goes down.” Really? Its as if I was speaking a different language. Krishnan Submaranian and I had the one thoughtful conversation I had at the conference on this topic. The topic of our conversation was“whose throat do you choak”, which he blogged on today. I couldn’t agree more with his post. Cloud and loosley coupled application developers will have to internalize these issues if they are to suceed in the enterprise and smb market. [..]
Dave,
I was talking about this.
[..] . In that post I had highlighted some of the important questions about data ownership, data portability, security, etc.. Few months back, I wrote another post aboutwhose throat to chokeif something goes wrong. In that post, I had highlighted the importance of holding cloud vendors responsible for downtimes. SLAs can answer these questions (and more) and, also, give a single point of contact to address any issues that might arise when enterprises move their apps and data to public clouds. In fact, Frank Ohlhorst has [..]