Aye and your local hard drive is any more reliable. Silly people.
As for Windows Azure going down...it didn't only the management dashboard because of certificate issues. Leap years got its claws in the certificate causing issues for connecting. But no data was lost. * Azure wasn't the only to have certificate issues on the 29th.
Perform regular backups and YES.....your hard drive is more reliable by far. Just because people consider backups as optional does not mean that HDD on pc is less robust than cloud storage......when you compare MTBF on pc HDD and raid configured HDD's on cloud servers you will find that the odd's of a failure is much higher for the server installation than on any single pc drive..........Read/Write IO is MUCH higher on a web facing server than on a personal pc. As for the fact that no Data was lost is mostly so much smoke and mirrors....if the cloud is not accessible then you are not storing data anyway so the database might as well have lost data.....you are still not doing business when this happens which is the REAL downside to cloud storage......management console or not, the data was not accessible which means anyone needing access to perform work was essentially up the creek without a paddle.....don't really understand what part of this people don't get when they talk about cloud storage.....it makes no difference what part is broken if you can't transact business.
Not really sure why this occurred when Microsoft SHOULD HAVE HAD fail-over standby installations ready at all times.....if not, then there is no reason that they should be in the cloud business in the first place........Are they a professional organization or not......failure to prepare for disaster recovery is very, very unprofessional, especially for a company like Microsoft......any monkey can keep from loosing data with SQLserver....the mark of true professionals is whether or not the database is accessible on a 24X7X365 basis....this is where redundant fail-over web/app servers really pay for themselves. It's not as if Microsoft has to pay licensing for the fail-over servers the way their customers do....this should have been built into the contracts with their customers.
As a Database Admin this failure really speaks to me about the level of or lack of professionalism that the customer receives for their money and should never have happened. Someone's head should be rolling.
There is no excuse good enough to excuse Microsoft for this type of failure, after all this is their business and they would be the first to point out this point of failure should someone contract with them for consultation on data center design and setup. This exact type of failure is elementary in Disaster Recovery planning.
My Computer
- Computer Manufacturer/Model Number
- Systemax N2000 Gaming PC
- OS
- Windows 7 Ultimate x64
- CPU
- Q6600 @2.4ghz (G0 stepping)
- Motherboard
- XFX nforce 680i LT
- Memory
- 8 gb OCZ vista essential sli PC-6400
- Graphics Card(s)
- Dual 9800gt in SLI mode
- Sound Card
- Integrated 8.1 High Definition Audio
- Monitor(s) Displays
- Dual Sceptre x246w 24 inch monitors
- Screen Resolution
- 1920 x 1200 each monitor
- Hard Drives
- 500 GB SATA II / 7500 rpm
- PSU
- Cooler Master Real Power Pro 1250W
- Case
- N2000 server tower
- Cooling
- Thermaltake Bigwater 760 is
- Keyboard
- MS Intelitype 6000 v2.0
- Mouse
- MS Intelipoint 6000
- Internet Speed
- Wi-power 1.5GB up / 512k down
- Other Info
- Windows 7 Ultimate x64, Windows 7 Pro x64, Windows Home Premium X86, Windows XP pro, Windows Home Server x86, Ubuntu 10.4 x86 and x64, Ubuntu server 10.4, SQL Server 2005, MySQL 5.0