So this is apparently something that we live with while the drive is brought up to date. I thought standalone RAID units usually supported an update-while-idle mode so that replaced drives are only updated when the controller isn't otherwise consuming all of the unit's bandwidth. But perhaps the unit is very busy or the drives are really slow or ...Eric the Sys Admin wrote:I apologize for any inconveniences you may be experiencing. The server your site is on had a hard drive go bad and it's been replaced. Since we use RAID 10 arrays on our machines, the new drive needs to be rebuilt into the array and this can take some time. During this process, you may notice slow load times or script timeouts. Once the rebuild completes, you should no longer notice these issues. Some more information is included below.
Then a paragraph of canned response text describing what a RAID unit is.
Our shared and reseller servers house data across multiple physical hard disks for speed and redundancy. The disks combined are called an array or logical device. When one of these physical disks fails, the server can still function because the data remains on other physical disks in the array as well. However, when a physical disk does fail, the array goes into a degraded mode so you will notice slower disk access times, etc. When we replace the failed physical disk, the data in the array needs to be rebuilt back on to the new physical disk to get the array back in an optimal functioning mode. This rebuild process is what takes time and disk access times are much slower than normal during this operation.