RAID member lost, OSX
Posted: 22 Mar 2013, 08:06
Hello, maybe you can help me with my problem. I sure hope so!
I have a 4.5Tb four disk RAID5 in an enclosure driven by a HighPoint RocketRAID 2314 card on a 2010 Mac Pro running OSX 10.6. Recently the volume spontaneously dismounted. At HighPoint's suggestion I deleted the RAID and set it up again without deleting data. This allowed me to recover about 200Gb of the material before it dismounted again.
High Point could not help me or tell me why my RAID was gone, but their WebGUI tool did tell me that two drives, #1 and #4, had too many bad blocks. I used dd to do a sector copy of both those drives onto two fresh ones. When I put them in the enclosure, the RAID began to rebuild itself. I let it run its course (possibly a mistake!) but the volume stll did not mount.
They recommended testdisk although they also freely admitted they had not used it and were going on previous customer recommendations. Let me say now that I am not a technician, I'm a video editor, so I am already quite confused by most of what is posted on this forum! But I have read through trying to understand...
Anyway, I physically loaded each dirve, one by one, into my Mac so I could use testdisk to check them. The reports for units 2,3 and 4 seem to be OK - I saw Mac HFS partitions as well as the volume name.
Unit #1 gave me the "disk too small" error. Here it is:
Disk /dev/disk3 - 1500 GB / 1397 GiB - 2930277168 sectors
The harddisk (1500 GB / 1397 GiB) seems too small! (< 6000 GB / 5588 GiB)
Check the harddisk size: HD jumpers settings, BIOS detection...
The following partitions can't be recovered:
Partition Start End Size in sectors
> Mac HFS 410336 8790082191 8789671856
Mac HFS 412792 8790084647 8789671856
Mac HFS 416432 8790088287 8789671856
Mac HFS 418120 8790089975 8789671856
Mac HFS 420528 8790092383 8789671856
Mac HFS 423616 8790095471 8789671856
Mac HFS 432296 8790104151 8789671856
Mac HFS 434440 8790106295 8789671856
Mac HFS 436984 8790108839 8789671856
Mac HFS 437024 8790108879 8789671856
It's the only one with that error - the rest seem to be completely fine - but then again, what would I know?
I've read about "adjusting the cylinders" when this error comes up, but I'll admit I'm at a loss to understand. Almost all of those threads are about single drives, not members of a RAID. My guess is that if I go to the "geometry" section and adjust them - then what? I'll be able to "see" the partition (which I can do already, btw) but then what do I do? This is a RAID5, so I'm assuming all my backup files are striped across the other drives. If I backed them up, what would I do with them then?
High Point seemed to think there was some way of "restoring the partition map" of the damaged member, and that this would allow me to recreate the array. But I'm over my head! Please, can anyone help?
I have a 4.5Tb four disk RAID5 in an enclosure driven by a HighPoint RocketRAID 2314 card on a 2010 Mac Pro running OSX 10.6. Recently the volume spontaneously dismounted. At HighPoint's suggestion I deleted the RAID and set it up again without deleting data. This allowed me to recover about 200Gb of the material before it dismounted again.
High Point could not help me or tell me why my RAID was gone, but their WebGUI tool did tell me that two drives, #1 and #4, had too many bad blocks. I used dd to do a sector copy of both those drives onto two fresh ones. When I put them in the enclosure, the RAID began to rebuild itself. I let it run its course (possibly a mistake!) but the volume stll did not mount.
They recommended testdisk although they also freely admitted they had not used it and were going on previous customer recommendations. Let me say now that I am not a technician, I'm a video editor, so I am already quite confused by most of what is posted on this forum! But I have read through trying to understand...
Anyway, I physically loaded each dirve, one by one, into my Mac so I could use testdisk to check them. The reports for units 2,3 and 4 seem to be OK - I saw Mac HFS partitions as well as the volume name.
Unit #1 gave me the "disk too small" error. Here it is:
Disk /dev/disk3 - 1500 GB / 1397 GiB - 2930277168 sectors
The harddisk (1500 GB / 1397 GiB) seems too small! (< 6000 GB / 5588 GiB)
Check the harddisk size: HD jumpers settings, BIOS detection...
The following partitions can't be recovered:
Partition Start End Size in sectors
> Mac HFS 410336 8790082191 8789671856
Mac HFS 412792 8790084647 8789671856
Mac HFS 416432 8790088287 8789671856
Mac HFS 418120 8790089975 8789671856
Mac HFS 420528 8790092383 8789671856
Mac HFS 423616 8790095471 8789671856
Mac HFS 432296 8790104151 8789671856
Mac HFS 434440 8790106295 8789671856
Mac HFS 436984 8790108839 8789671856
Mac HFS 437024 8790108879 8789671856
It's the only one with that error - the rest seem to be completely fine - but then again, what would I know?
I've read about "adjusting the cylinders" when this error comes up, but I'll admit I'm at a loss to understand. Almost all of those threads are about single drives, not members of a RAID. My guess is that if I go to the "geometry" section and adjust them - then what? I'll be able to "see" the partition (which I can do already, btw) but then what do I do? This is a RAID5, so I'm assuming all my backup files are striped across the other drives. If I backed them up, what would I do with them then?
High Point seemed to think there was some way of "restoring the partition map" of the damaged member, and that this would allow me to recreate the array. But I'm over my head! Please, can anyone help?