Frimware Upgrade

[choujij]

Hi.

I was transferring a file and recently received an error from Amd Raid Xpert that mentioned a timeout on one of my PC's Raid volume. This volume is Seagate 2TB Green drives running in RAID0. I read on the net something about a firmware upgrade that Seagate released to eliminate some bugs with AMD's raid software. So I was just wondering:

1) Can I just go into my bios settings and disable my raid volume, upgrade my drive's firmwares, and then re-enable the raid volume without it affecting the data in my drives/raid array? I know it's always risky to do so, but under normal circumstances I think this should work just fine. I just wanted to know if there is a reason this will not work?

 

[hang-the-9]

Running RAID 0, I take it you have backups? If not, get some. You should be able to update the firmware without disabling RAID. Seagate would have info on that I'm thinking.

 

[choujij]

No, I don't have a backup and yes I am aware of the risks. I've never run into any trouble in the past with matched Seagates in RAID 0. The 7200.12 series have treated me well, and work really fast in RAID 0. That said, if I ever do have one fail, I'll just purchase another identical drive and swap out the platters (which I've done in the past). One time I even visited Seagate during their colossal 7200.11 failure period, and talked to their lab-coat wearing techs about the process.

Anyways, back to the point. It turns out the Seagate Green Drives are not really compatible for running in raid due to the eco-software programed in the firmware. Apparently, the RAID software does not know how to work with these types of firmwares properly (unexpected responses, spin-up times, etc.). So I am going to remove my data from the drives and disable the raid array.

Seagate does not have a firmware to address these kinds of glitches, because their view is that these drives are not intended to be used in RAID. =S So in my particular case, what was happening was that once the drive was 1.8TB full (the volume held 3.6TB), the drives stopped writing any further data. I could re-write what was already on the drive, but could not exceed 1.8TB of data. I'm almost inclined to believe their was something wrong with the partition table, but given the other errors I'm likely to incur, I'll give up at this point.

 

[hang-the-9]

For Raid, use WD Black or Samsung Spinpoints. Green drives only useful purpose is to save energy, although with how little that savings actually is, unless you are running a hundred of them 24/7, they are a bit of a waste.

 

[fzabkar]

AIUI, Seagate's firmware updates cannot be applied to drives that are behind a RAID controller. The SeaFlash utility won't detect them.

Note the warnings in the following document:
http://seagate.custkb.com/seagate/crm/selfservice/search.jsp?DocId=207963

"Caution: Customers utilizing RAID architectures incorporating affected products should back up all data content before upgrading drives with the firmware solution. For some arrays firmware upgrades may cause RAID management software to require an array to be rebuilt. Please consult the technical documentation and procedure recommended by the RAID controller manufacturer for your system prior to upgrading any affected product in an array."

"This firmware update procedure is for SATA drives attached to standard SATA controllers. The download utility, SeaFlash, may not see drives attached to RAID adapters and you may need to move the drives temporarily to a standard SATA controller to run the firmware update procedure. Intel RAID configurations are supported and tested through ICH10R. Please see Seagate KB article: Firmware tools for SATA RAID adapters for additional information."

Firmware tools for SATA RAID adapters [212717]:
http://seagate.custkb.com/seagate/crm/selfservice/search.jsp?DocId=212717&NewLang=en

 

[clarkjd]

So in my particular case, what was happening was that once the drive was 1.8TB full (the volume held 3.6TB), the drives stopped writing any further data. I could re-write what was already on the drive, but could not exceed 1.8TB of data. I'm almost inclined to believe their was something wrong with the partition table, but given the other errors I'm likely to incur, I'll give up at this point.

Are you sure you weren't running into the 2TB limit of drives using the MBR?

 

[choujij]

Are you sure you weren't running into the 2TB limit of drives using the MBR?

Yes, both drive were formatted with GPT instead of MBR.


So I removed my data from the drives and disabled the RAID array in my bios. I then went into Windows, and converted each of the drives to dynamic disks. From there, I created a striped volume that spans over both drives. This essentially allows me to have both drives acting as one again, striped, without utilizing the RAID controller.

However, the same thing happened again, only this time with 1.91TB free remaining. I started transferring four 1GB archives, which began the process at 240MB/sec, but then after 20 seconds it slowed down to a rate of 20MB/sec. So frustrated, I canceled the transfer. I then tried transferring 1 archive at a time, and each went through much faster (around 73MB/sec). After transferring 10 more GBs of data, the speed of the drives went back up to 240MB/Sec.

I'm not sure what to make of this.