It appears that the adage, “When it rains it pours,” applies when it comes to life in the clouds. Going in reverse order, three giants of cloud computing have had recent issues…
From InfoWorld today:
Punch line about Microsoft’s Business Productivity Online Suite (BPOS)?
If I count correctly, BPOS mail service went down on Tuesday, with some customers out for six to nine hours. It went down again at 9 a.m. on Thursday and came back up at 10 a.m. with a 45-minute additional delay. It went down again at 11:30 on Thursday, was out for half an hour, but in that time, 1.5 million messages got backed up, and they still hadn’t all cleared by 4 p.m. Thursday — not a good week for BPOS email.
And the piece is even more damning about Microsoft’s non-communication during all of it.
Well, that’s the “old horse” Microsoft…nobody expects much out of them. Luckily Google has their act together…except…err…that issue with a Blogger outage that, at least temporarily, also whacked a couple of days worth of posts:
Nothing like a wayward maintenance release to gum the works…
That ZDNet piece rightly notes:
Google has owned and operated Blogger since 2003. It’s not like they’re still trying to figure out how to integrate the service into their operation. If it can happen at Blogger, why can’t it happen with another Google service?
Finally, most famously (or infamously?) and with the most impact to companies affected…Amazon’s failure:
Nobody wants to have CNN use phrases like “massive crash” when discussing their service…and you’ll note again some discussion of a major corporation being tight-lipped (at least until afterward when they drowned people with a wordy explanation)…along with maintenance gone wrong (an upgrade, more specifically). Perhaps most concerning is that after being down a few days:
Amazon said a very small percentage of data storage volumes in the east region of its cloud service are not “fully recoverable” following the lengthy outage that started last Thursday. As of Monday, Amazon’s cloud-based web services were up and running for most customers, but the company said about 0.07 percent of the data stored in the east region could not be recovered.
Ouch. That TechFlash article (“Amazon: Some data lost in cloud outage is not recoverable”) will also bring back some bad memories for those of us who have been at the mercy of a third-party provider whose outage is killing us, but who is also giving us no clue when we’ll be back up:
“To be told four to six hours, and then just wait and wait and wait and nothing’s happening for three days, was beyond frustrating and maddening,” [Crosscut.com associate editor Michele Matassa Flore] said.
So, does this mean that we should avoid cloud computing? No…local resources can fat-finger something as quickly as third-party ones. However, business continuation plans should not assume perfection out of anyone, and realize that your importance to any provider is based on how big you are compared to their myriad other clients (just listen to Flore in the TechFlash post).
Basically, as long as it is likely a third-party will be more reliable than you will be, that they’ll provide better service, and that they are cheaper…head for the clouds…but always prepare for the worst no matter how big a name.
There is a solution…