Broken HDD in RAID 5

[xBlizzDevious]

I was just playing a game and I suddenly heard a deafening screech. Turns out it was the buzzer on my RAID card (HighPoint RocketRAID 2720SGL - as recommended by someone on this forum) and one of my HDDs is faulty.

 

It says it's the drive on port 4; which I have labelled so I know which one it is.

 

Basically, what I want to know is; what is the process of replacing a drive in a RAID 5?

 

I don't have any replacement just yet, I'll need to RMA first.

[Eric1024]

The basic steps are:

 

-Power down system

-Remove bad drive

-Add new drive

-Reboot into the BIOS of the raid card or boot into your OS and use whatever software that HighPoint has for raid management

-Find the option to rebuild the array

 

I know that's a bit vague, but I don't have direct experience with HighPoint cards. Also, look on their website and look for the documentation for your card. Should help.

[xBlizzDevious]

That sounds exactly as I was expecting. I haven't removed my HDD yet as my RAID is working just fine now - though one drive down - and I will have to RMA before I get the drive back so it may be a couple of weeks on 3 drives.

 

Thanks for replying! I've come to expect answers within the next hour on this forum! Haha. Was surprised when I hadn't got a reply a day later.

[dalekphalm]

That sounds exactly as I was expecting. I haven't removed my HDD yet as my RAID is working just fine now - though one drive down - and I will have to RMA before I get the drive back so it may be a couple of weeks on 3 drives.

 

Thanks for replying! I've come to expect answers within the next hour on this forum! Haha. Was surprised when I hadn't got a reply a day later.

Keep in mind that if you continue to use your degraded RAID 5 array - for up to three weeks no less - you run the risk of losing your entire array and all data on it. If a second drive fails before you get that replacement, then you're boned.

 

Try to keep use to the bare minimum until you get the replacement in.

[xBlizzDevious]

Keep in mind that if you continue to use your degraded RAID 5 array - for up to three weeks no less - you run the risk of losing your entire array and all data on it. If a second drive fails before you get that replacement, then you're boned.

 

Try to keep use to the bare minimum until you get the replacement in.

 

Cheers. You got it. Thing is, I'm an SSD with OS and HDDs - or more specifically in my case, RAID 5 HDDs - for everything else guy. I don't actually have much truly important stuff on them and what I do, is backed up to the cloud.

 

The main reason I fear these drives failing is the time it will take to re-download the many terabytes of data that's on them with my 5mbps down connection......

[IdeaStormer]

That sounds exactly as I was expecting. I haven't removed my HDD yet as my RAID is working just fine now - though one drive down - and I will have to RMA before I get the drive back so it may be a couple of weeks on 3 drives.

 

Thanks for replying! I've come to expect answers within the next hour on this forum! Haha. Was surprised when I hadn't got a reply a day later.

 

You still might want to buy a second drive as a backup, as after one fails the rebuild process sometimes causes a second or more other drives to fail. This is of more importance if all the drives are identical and of the same age, read up on this and see for yourself.

[xBlizzDevious]

You still might want to buy a second drive as a backup, as after one fails the rebuild process sometimes causes a second or more other drives to fail. This is of more importance if all the drives are identical and of the same age, read up on this and see for yourself.

 

Well they're all the same make and model and they're also all only a few months old....

[dalekphalm]

Well they're all the same make and model and they're also all only a few months old....

I believe what he is referring to is if all the drives are from the same manufacturing batch, then the entire batch could be defective or susceptible to premature failure. It does happen from time to time. So if one dies prematurely through no fault of your own, the risk of all of them failing within a similar time frame is significantly higher then say if you had bought the drives at different times or from different retailers (Thus ensuring that they wouldn't be from the same batch).

[TheProfosist]

The basic steps are:

 

-Power down system

-Remove bad drive

-Add new drive

-Reboot into the BIOS of the raid card or boot into your OS and use whatever software that HighPoint has for raid management

-Find the option to rebuild the array

 

I know that's a bit vague, but I don't have direct experience with HighPoint cards. Also, look on their website and look for the documentation for your card. Should help.

Yep this would be the way to do it and the way ive done it countless times.

You still might want to buy a second drive as a backup, as after one fails the rebuild process sometimes causes a second or more other drives to fail. This is of more importance if all the drives are identical and of the same age, read up on this and see for yourself.

I also agree you should always have an extra drive on hand. if you need 4 buy 5, that way when one drops you just pop in the new one youve got siting on the shelf and rebuild. The you send off the bad one to get RMA'd, and when that one comes back its your extra. I learned this the hard way in the past...

[TheProfosist]

I believe what he is referring to is if all the drives are from the same manufacturing batch, then the entire batch could be defective or susceptible to premature failure. It does happen from time to time. So if one dies prematurely through no fault of your own, the risk of all of them failing within a similar time frame is significantly higher then say if you had bought the drives at different times or from different retailers (Thus ensuring that they wouldn't be from the same batch).

Agreed but for home use or even small business use its nearly impossible or stupidly hard to do that. Large enterprises do this so that if there was an issue with a batch it doesnt compromise everything.