>Another way to bust your data is to rely on RAID 5 or 6 or similar, at >least if the software does not take special care. > >When those mechanisms, updating a block always results in at least 2 disk >writes: The data block and the checksum block. There's a small time >window where only one of those blocks physically reached the disk. Now, when >the >power fails during said time window, and the third disk fails, it's >content will be restored using the new data block and the old checksum (or vice >versa), leaving your data garbled.
Generally this is only an issue with fake-RAID (aka software RAID). Hardware RAID will issue the writes to update the stripe in parallel across all spindles which need to be updated. Moreover, although writes to a hardware RAID device are signaled complete once the data has been written into the buffer on the RAID controller, the hardware will take special precautions to ensure that any write which makes it into the hardware buffers is properly written to disk even if there is a power failure before the scatter-write-with-verify to the physical media has returned completion-without-error for all spindles. You will only lose data if the power is out for longer than the battery on the hardware controller can maintain the buffer -- and the better classes of hardware raid contains NVRAM to which "dirty" stripes are flushed on power loss so that they can written to the physical spindles even if the power is not restored until long after the buffer RAM battery has lost power. In other words, you get what you pay for. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
