Heh, just a small gasp of air and a quick note where attentions have been.
As mentioned in a previous post “Servers, Drivers and Time – Oh My!” a main server at work “caught the flu” and has been giving grief.
The first wave of problems and repairs were followed by a second, and then a third. Each more serious.
Much of the repairs then needed ended up a team effort between myself, the hardware specialist and the network specialists. In the end the machine was physically gone through, numerous components replaced and the server rebuilt.
Tape backups let us recover to a reasonably close date, as the combined hardware/software “melt down” did spoil the last working day’s data entry & transactions. Seems it reported to the backup software that “No Files Have Changed” and claimed “Successful Backup Complete” after a multi-hour job had run but a moment or two.
The failure sequence is interesting, as devices intended to reduce the risk of failure were the problems. The backplane of the server’s RAID physical server array failed, taking out one of the five drives completely while damaging the data on the remaining enough to negate the advantages of Mirroring, Striping and Stand-By Hot Swap drives.
Once the backplane was replaced it was another 12 hours until the Host Drive failed and the magnitude of the data damage to the other drives became apparent.
In condensed version, with much help the unit was physically repaired, the RAID array rebuilt, Data restored from tape and then all the active directory & SQL database nuances repaired.
We’ve also started a replacement server build, as I have little faith in hardware that suddenly shows an unexplained series of failures.
The ongoing epiphenomenon might be ended, but as there is no way to predict if this sequence of failures will not restart, the server gets replaced.
Technology moves along so rapidly that the update will take this server from a Single Pentium Server Processor to dual Quad-Core Xenon Processors. The RAID array jumps both in size to about 1/2 TB but greatly in read-write & data throughput speeds.
In the midst time was made for a visiting Basque friend & his family, a family reunion, and getting Tom KC9JGD off to Freshman Orientation at Michigan Technological University 4-1/2 hours away.
More one each of these and release of some of the draft articles I just didn’t have time to finish in the next weeks.