I am reviewing some redundancy based checks for a database application (MSSQL) but I wanted to gather your view as to whether these best practices are more geared towards physical servers and opposed virtual servers running on a SAN.
The best practice suggests separating database and log files and backups onto different drives, indicating if one drive dies then the other should be ok and it may not be a disaster.
But for drives on a VM hosted on a SAN - can individual drives on the VM still fail as they could in old physical drives. And by fail I mean can they fail in isolation, or would it be more likely the whole VM would fail as all the "virtual" drives on the VM are on the same hardware (SAN). If virtual drives can still die/fail in isolation even though they arent unique hardware as they would have been on an old physical server, can you give some pointers on how/common reasons they would fail/crash/whatever?