Mau1wurf1977 wrote:Yea I stuffed up. Meant the mirrored raid of course 🤣
😁
I remember those old days.. I've been very meticulous when designing storage configuration. When the storage comes with many hard drives, I usually create separate RAID volume for index files, separate RAID volume for redologs (rollback segments), separate RAID volume for log files, separate RAID volume for reference tables, separate RAID volume for transaction tables.... etcetera. Just because I want to make sure that everything runs on different hard drive spindles to maximize performance.
And the systems engineer said I'm an obsessive-compulsive database admin.
Those days are over now. SAN vendors keep claiming that hard drives are getting faster and faster that different spindles don't matter anymore. These days, SAN vendors like NetApp and Fujitsu are putting more emphasis on automation and easy to use than control. When you create RAID volume, it is now virtual volume, and you cannot specify which physical hard drives belong to which volume anymore.
Guess it doesn't work well for control freaks like me, but that's just the way it is, I think. At first, it was hard for me to accept the fact that I cannot put different database files on different spindle. But well, time changes.
And things get even "worse" with the arrival of virtual machines.... You know, VMWare and such. Database servers no longer have separate physical volumes for storing various data files, since even the hard disks are VMDK files.