Ballmer Says Obvious: SideKick Outage "Not Good"
Here's a little update for all you Sidekickers who lost data over the last couple of weeks. In a statement released yesterday, Microsoft assured users it is working "around the clock" to restore everything.
In a statement released yesterday, Microsoft assured users it is working "around the clock" to restore everything. With any luck, affected users should have their contacts back by the end of this week.
"We continue to make steady progress, and we hope to be able to begin restoring personal contacts for affected users this week, with the remainder of the content (photographs, notes, to-do-lists, marketplace data, and high scores) shortly thereafter."
Nice to see everyone is going to get his or her important data back, even if it is going to take a while. That said, what would this cloud computing catastrophe be without a word or two from Steve Ballmer?
When asked about the incident, he described the situation as, "not good" and went on to say that it looks like all the data will be recovered. The CEO of Microsoft went on to tell Network World that the Redmond-based company would have to make extra effort to assure to enterprise customers the same mistake would not be made again.
"It is something we are going to have to address and explain to customers our method and process and quality approach and what went wrong in that case and how we are making sure that it does not happen again."
Would you ever trust cloud computing? Why or why not?
- Microsoft Store Opening With Win 7, Ashley Tisdale
- Modern Warfare 2 PC: No Dedicated Servers
- Marvell Launches Fleet of Armada ARM CPUs
- Dell Adamo XPS's Latch is Heat Sensitive
- Sigh of Relief: Modern Warfare 2 PC Not Delayed
- VIDEO: Sneak Peek at Win 7 Family Guy Show
- Intel to Ship Value-Based X25-X SSD in Q4
- Hitachi's New HDDs A Blazing 15,000 RPM Fast
- Samsung OLED is Virtually Indestructible
- New Triple Core Athlon IIs Are Great Value CPUs
- Apple's White MacBook Goes Unibody Too
- Apple's New Mouse is Magic, Has Multitouch
- Apple iMac Goes Core i5, i7; Mac Mini Updated
- Microsoft Releases Sidekick Data Recovery Tool
- Time Warner Cable's 65,000 Routers Open to Hack
- DISASSEMBLED: Apple's Polycarbonate MacBook
- Sun Microsystems Cuts 3,000 Staff Due to Delays
- ''Pipe Dreams'' Computer Desk System, $30,000
I would if it was across 6 to 14 data centres, each at least 250 km apart, with some form of delayed replication. Or delayed and 'instant' replication done in an intelligent way.
That way if some amateur breaks something, the others have time to step in and fix this sort of thing from happening again.
Hopefully a giant like Microsoft (and Danger Inc, Sidekick, Hiptop, Telstra, etc) will learn from this and actually implement redundancy across multiple data centre sites.
It only takes one bad line of code, and bam, with high levels of automation one could destroy an entire data center (logically, not physically, unless robot trains run amok and break stuff too).
ie: Anything 'destructive' (from the user land perspective) should be buffered / delayed transactions to/from other data centre sites.
God only knows how they managed to screw up so badly, thankfully I switched to an iPhone 3G series handset from a HipTop3 over a year ago.
sure i suppose "trusting cloud computing" in general is part of the fallout here -- if a label like "cloud computing" even has to take the blame.
in this example here i think of the issue as more a Microsoft issue and would have to ask users affected by this data loss if they can trust with their data to Microsoft again -- how Microsoft is keeping the data safe, i dont think their customers care, they simply want to count on it always being there. trust can be broken by a single incident like this and cannot be rebuilt again overnight. it'll go a long way toward rebuilding if Microsoft is more transparent about how it /used/ to handle data and how it will handle data from now on. they could use this issue to even set a new standard for themselves, and perhaps for their competitors. but of course you cant just say it, you have to do it and prove that youre doing it, to rebuild trust.
and with regard to a business using cloud computing, i'd say its really not a new problem to solve. with data, like with security, internal standards are important, policy is important layers (at the technical level) are very important and so is redundancy even when it comes to cloud computing.
if the data is important, and to most businesses it is /the/ most important thing. take care of it by planning for very bad things to happen. having policies in place for architecting a new solution and auditing it once its up and running ... even if youre contracting it out.
again, easy to say but hard to do, especially when your whole team can get distracted by just trying to hit its dates for project managers.