Difference between version and version      View first change»»

Back to News, or News version history

At line 9 added 11 lines.
!!25-Apr-2012 Sensitive little servers...
Tonight at the Melbourne DC there was some work done on the air conditioning system.
Two of our servers alerted us to deviation in temperature, and two more alerted us to some strange voltage fluctuations. We contacted the M2/Primus DC and asked what was going on and were told about the reset of the air conditioning. No other customers of the DC had reported any issues, so I guess we're monitoring more than other customers. We had 4 servers reporting issues, so it raised a definite red-flag with us. 20-30 minutes later 2 of the servers told us everything was ok again with voltages, and 10 minutes after that the remaining servers reported A-OK on the temperature variance.
Interestingly, during the same period, drive speeds on those servers dropped dramatically down from 240MB/s to around 3-4MB/s. I haven't seen that before and it was a strange occurrence. If drive speeds reduce dramatically on a server it usually indicates a RAID-battery failure or RAID-array problem. Two servers at the same time though? And it happened to be the same two which had alerted us about voltage fluctuations. After the alert-period finished and everything was ok, drive speeds immediately returned to their normal rate - literally within 1 second - I was watching them at the time while working in another window trying to determine the reason for the degradation. I can't find anything in the Dell documentation to explain why that would happen, but I guess the servers were just being extra careful.
Sensitive little beasts, but we love our Dells.
metawerx specific

referring pages