Sidekick's lost data gone for good
October 12, 2009 by Lin Edwards
(PhysOrg.com) -- Sidekick users have been without some of their services for days, and have just been told by the company, T-Mobile, that for some users their data may be lost forever due to a server error at Microsoft subsidiary, Danger.
T-Mobile has published an apology to affected users for what is turning out to be one of the most spectacular failures in cloud computing. The massive data loss also suggests entrusting personal data to the cloud may not be as safe as users have been led to believe. The loss of services and data was caused by a disruption to services, but details are sketchy.
The problems began last week with a massive outage at Danger. Many Sidekick users lost access to their key data, because almost all the data is stored in the cloud, with no copy on the device itself. On Saturday, Microsoft admitted that any data it had not yet recovered is almost certainly permanently lost.
T-Mobile has offered to compensate affected users with a free month of services, but neither T-Mobile, nor Microsoft have revealed how many users have been affected.
Users who have lost their data would need to rebuild it themselves, but many have not backed up their data since they believed it was safe. Data affected includes contact lists, calendar details, personal information and photographs.
T-Mobile is warning users not to power off their Sidekick device, to reset it, or to remove the battery or allow the battery to go flat. If they do, they will lose any remaining personal data they currently have in their device, since the data that was backed up on the Danger servers has been lost.
Another update from T-Mobile and Microsoft on the current state of the repairs is expected to be issued today.
© 2009 PhysOrg.com
Feb 17, 2009 |
not rated yet |
0
Oh yeah, I'll be rushing right back to them for that free month of secure (cough) service.
Can you hear that? It is someone singing the azure blues. :-(
Many have speculated about the security of cloud computing systems and the volatility of data placed there. This makes all those conspiracies seem plausable.
I agree with CreepyD, we should be informed what DataCentre problems caused not only the cascading failure of the servers in the farm/cloud, but also to the loss of data that should exist on nearline and offline backups.
If this was a hack, then the infiltrators gained access to multiple sets of data, (live and backups), or corrupted the backups for a significant period before attacking the live data.
If this was NOT a hack ... the live data and the backed up data were corrupted and/or lost by the cloud database itself, (and the staff that manage it), which is of course the worse scenario.
Mobile phones have small datasets. What if they were big payroll systems?
So just make sure your mobile device has an infinite battery capacity and you'll be fine! Good old Microsoft problem solving.