Migrated from eDJGroupInc.com. Author: Greg Buckles. Published: 2012-11-14 04:10:45Format, images and links may no longer function correctly. Migration to the Cloud is one of the key trends that eDJ Group is calling out for 2013. Almost every major corporate and law firm eDJ client in 2012 either already has data in Cloud services or is exploring the options. As CTO of the eDJ Group, I finally retired my Exchange box and put the eDJ Group on MS Office 365 last year. The potential savings are undeniable, but what happens when your legal department is under the gun to collect from these sources and they go offline? Every one of those Cloud providers will talk about 99.9% uptime guarantees and most will deliver. At the recent Masters Conference, my friend, who also happened to be keynote, Google’s Jack Halprin, pointed out that eight hours of downtime in a year is really just a drop in the bucket. He was right, but that depends on how you define downtime and when they occur. It appeared that while preparing to roll out service updates, Office Web Apps and more, North American users of Office 365 suffered intermittent connectivity and accessibility errors for the second time in 5 days. I know that it happened because my phone was ringing as my remote teammates tried to reset passwords, reload Outlook and frantically ran virus scans thinking that the problem was on their end.
Microsoft’s lack of response to support requests set off a flurry of Tweets, forum posts and web traffic starting as early as 7 AM EST. For anyone who is considering the migration of their vital communications to Office 365, take a read to the responses to Microsoft’s official forum post about a “potential issue in North America” at 11:24 AM. One way to keep from having to pay downtime guarantee rebates is to just not recognize an outage. After all, does an outage have to affect the global Office 365 system, regional data centers, specific servers or just specific accounts to be reported and tracked? I know that Babs Deacon in New Jersey was offline very early yesteday morning, yet the first word of investigations did not show up on our status dashboard until 11 AM CST. Worse, the outage on November 8 does not appear in our status dashboard because we did not go offline. The Office 365 team finally tweeted that the outage was resolved at 6pm EST. So how do you really count downtime? The last thing that anyone under a production deadline wants is to be helplessly left offline with no acknowledgement, insight or options into the problem.
However, I am not painting Google or any cloud provider with the Microsoft support brush. I am simply pointing out that putting eDiscovery eggs in the care of global cloud providers requires a completely different kind of risk assessment and disaster recovery plan. If Amazon S3 is down, then a surprising proportion of eDiscovery hosting providers are probably down as well. Taking a lesson from Microsoft, Amazon’s status dashboard makes it very difficult to see historical outages. However, it does not take a Google genius to find outage articles and posts. Do Cloud providers suffer more downtime? Not that I can tell. In fact, they probably have a much better record than the vast majority of corporate IT or eDiscovery provider data centers. However, when those go down during critical times, you have a throat to choke and an escalation path to find a speedy response. So before you uplift or outsource all of your ESI and eDiscovery systems, remember that you still need a disaster response plan for outages, even if you do not own the Cloud. You cannot abdicate your responsibilities and point the court at Microsoft, Amazon or your provider when you miss a deadline. Ask the hard questions, have a plan and measure your risk before you accept it. The cloud is here and you have ESI in it already, so it is time to start including it in your disaster and eDiscovery planning.
eDiscoveryJournal Contributor and eDJ Group Lead Analyst – Greg Buckles