RAID Failure question

[PaxtonTheSpy]

Hi,

I've been looking all over for an answer to this problem, but haven't found anything yet. Hopefully you guys can help out a little.

I'm running 3 Areca 8050's filled with 8 4TB drives apiece (I store a lot of 4k video) all running off a Mac Mini Server running OSX 10.9.1 in a RAID 50. My newest array has failed, I believe it is a drive issue, but haven't been able to access the arrays status, which is where we get into the issue.

The array starts up perfectly fine, you get about 1-4 minutes of access before it dives into what seems to be an endless access loop. The RAID utilities are all network accessed and take a second to get to. I haven't been able to get all the way into the SMART status' before the access loop begins.

I have contacted the retailer and manufacturer. They sent out a replacement unit and the same issue is happening which is one of the reasons that I'm leaning toward drive failure. That and the drives don't have TLER (I have all HGST HUS724040ALE640 drives)

I've heard that it can be detrimental to take the slices attach a bare drive>USB connection and mount them on either my windows machine or the mac to check the drives, that they will lose some part of their array config. and will then not be able to be replaced in the RAID.

Do any of you have any ideas on how to check which drive may be failing or if it is a drive issue without being able to check the raid health through the manufacturers status interface?

I've been banging my head against my keyboard for a couple weeks trying to find posts.

I appreciate the help!

 

[Alec Mowat]

Anything attached to the device via USB, like an external drive?

 

[PaxtonTheSpy]

The three arrays are all attached in a Thunderbolt chain to the Mac Mini. There are peripherals attached to the Mini (keyboard, mouse, a couple flash drives occasionally)

 

[FireWire2]

Don't do ANY writing to the ANY volume...
Connect ONE box at a time to your TB port and run the mRAID
Log in look at the events log, you can determine which box is bad...
Once you have the BAD box power it ON - HOT REMOVE a HDD.
The alarm will goes off wait for 30sec then reinserted it back in
It will rebuild, the rebuilt process will mapped ANY bad sector on the HDD

 

[Alec Mowat]

The three arrays are all attached in a Thunderbolt chain to the Mac Mini. There are peripherals attached to the Mini (keyboard, mouse, a couple flash drives occasionally)

If the SMART isn't immediately picking up any errors, and you are using a new device, I imagine it's one of these daisy-chained drives that's causing the issue.

Disconnect everything and work with the NAS on it's own. See if it crashes in that case,

If it does, remove one drive at a time and boot it. Test with each drive removed and see where it goes.

 

[PaxtonTheSpy]

Don't do ANY writing to the ANY volume...
Connect ONE box at a time to your TB port and run the mRAID
Log in look at the events log, you can determine which box is bad...
Once you have the BAD box power it ON - HOT REMOVE a HDD.
The alarm will goes off wait for 30sec then reinserted it back in
It will rebuild, the rebuilt process will mapped ANY bad sector on the HDD

I'll try this in the morning, what will happen if one drive is completely toast and there is an attempt to reinsert into the array?

 

[PaxtonTheSpy]

If it does, remove one drive at a time and boot it. Test with each drive removed and see where it goes.

The NAS still crashes when booted on it's own and I still can't get in to check the SMART status on any of the drives.

You're the second to suggest removing drives. You are suggesting removing the drives while the array is powered off, starting the array, and seeing what happens minus a slice? What will be the difference between this and the other comment of removing a slice while the array is powered and running? I just want to have a complete picture before doing something that could potentially rebuild the array or possibly lose data.

Thank you for your help!

 

[FireWire2]

Here is how I would trouble shoot your issue.
What I forgot to tell you is connect to another MAC with TB, and check each RADI5 with mRAID.
Do not connect to Mac Mini Server . It wont' help to trouble shoot

 

[PaxtonTheSpy]

connect to another MAC with TB, and check each RADI5 with mRAID.
Do not connect to Mac Mini Server . It wont' help to trouble shoot

Tried this with a Mac Book Air, can you explain a little more about mRAID, I'm not familiar. Everything I've read has said to only use the manufacturers RAID health monitors, but this box starts an endless access loop that freezes all interfaces with the health monitor or even Finder in under 2min making it all but impossible to even get into the logs or the SMART status'.

 

[popatim]

Did you not setup email alerts?

You can use the lcd panel top check for some problems as well as by attaching to it via lan and using your browser.

 

[Alec Mowat]

If it does, remove one drive at a time and boot it. Test with each drive removed and see where it goes.

The NAS still crashes when booted on it's own and I still can't get in to check the SMART status on any of the drives.

You're the second to suggest removing drives. You are suggesting removing the drives while the array is powered off, starting the array, and seeing what happens minus a slice? What will be the difference between this and the other comment of removing a slice while the array is powered and running? I just want to have a complete picture before doing something that could potentially rebuild the array or possibly lose data.

Thank you for your help!

Power off. Remove a drive. Power back on. Test. If it works, it's probably that drive you pulled.
If it crashes, power off, replace that drive, pull another drive, power back on. Repeat until all drives are tested

 

[FireWire2]

The worst thing you can do the RAID5/6 is power off and remove a drive(s). Don't do that!

Connect to MB, connect a network cable to your RAID, then open a browser and log in with its IP address...
Unless you changed the log-in credential, use admin <-- login name and 0000 <-- pw
Check the status of the RAID and look through events log

Or load mRAID ( download from Areca site) and run the ARCHTTP <--- http proxy so you can access the RAID GUI. This you don't need network cable