Well, if one of the hard drives in a RAID 1 array was just a little flaky, then it is very likely that the shock shutdown may kill it.
Or, if there were some sort of OS corruptions hanging in the balance, they may show themselves with a no boot situation.
How do we know that one of the drives is flaky after the power outage?
The OS will take an inordinately long time to boot, and may, or may not, have a service failed error after the outage.
So, in this scenario we can discover very quickly two of our weakest links:
- UPS is not big enough to gracefully shutdown the server after a predetermined amount of time
- Our backup restoration method has not been tested yet and it fails.
So, we learn a very valuable lesson:
- Have a UPS in place that is big enough to allow for a graceful server shutdown
- Have a UPS or UPSs in place that properly filters all A/C power or other power sources heading into the server including:
- Telephone for fax and/or DSL
- COAX for cable Internet
- CAT 5E for all server ports
- Network switches, DSL/cable modems, etc protected.
- Workstations have at least a 3K+ Joule surge arrest in place on their A/C, CAT 5e, and telephone if needed.
- Workstations used for database work have at least a 1000VA UPS with filtration in place for graceful shutdown of the database.
Sometimes a factory defect may not show itself until the UPS flips over to battery mode for an extended duration. So, it is better to know that the unit is going to fail before an actual power event brings a defect out and causes even more grief.
Philip Elder
MPECS Inc.
Microsoft Small Business Specialists
*All Mac on SBS posts are posted on our in-house iMac via the Safari Web browser.
No comments:
Post a Comment