Proliant ML310 RAID problem

[Martin Briggs]

We have a Proliant ML310 with onboard IDE RAID controller. The server was originally setup with a raid 1 configuration but at the begining of the week the server fell over and when restarted stated that the RAID array was degraded. The server ran on one drive for a couple of days and then HP asked us to do a restart to confirm the error message which we did. Following the restart the server rebooted using the second drive that was of course two days out of sync. When we checked in the raid configuration utility we found a RAID 1 array (showing as degraded) which was the drive that it had been booting from up until the reboot and also the second drive (that it is now booting from) is showing up as RAID 0.
When the HP engineer arrived he was going to simply add a new hard drive in until we advised him that this was the second time this had happened in the servers 5 months of life and that when the first engineer had done this 2 months ago he had wiped the operating system of which meant we had to reinstall everything from scratch!! Because of this the engineer spent many hours looking into the problem and talking to his help desk and concluded that he couldnt understand how the system had failed into this state and that any attemp to add in a drive and recreate the array would end up wiping the system clean.
Thus the crux of the question is can we take a RAID 0 drive and then convert it to RAID 1 and then add a second drive that it can use to mirror its info too, all without wiping the existing data?
Many thanks
Martin

[Gert Luyten]

With the ATA Raid controller being a low bduget controller, there is no way of doing an online migrating from RAID0 to RAID1 without losing all data.