[SOLVED]Too many wrong or missing disks! Adaptec RAID 71605E


Recommended Posts

First let me start by saying that all my disks work fine in unRAID 5.0.5, so there's nothing wrong with my disks/cables or sata card. So I decided I was gonna try the v6.0 beta 7, when I first booted up web UI and syslog says I'm missing 2 disks, they are not showing up with serial number in the dropdown menu or the syslog. I rebooted then I was suddenly missing 5 disks, rebooted again and now missing 7 disks! Instead of rebooting I now did a powerdown, when I was booting up again I was missing 1 disk, did a powerdown again and and still missing 1 disk. Same thing happend when I was trying to boot xen/unraid. So I gave up and went back to 5.0.5, but clearly something is wrong when all my disks are working in 5.0.5 and not in the beta, I've attached the syslog from the last time I was booting up.

 

My setup is:

Motherboard: ASUS Z97-DELUXE

CPU:  Intel Core i3-4360

RAM: 16GB Crucial DDR3 BallistiX Sport 1600 Mhz

RAID/SATA controller: ADAPTEC RAID 71605E SGL/256 SATA/SAS

Disks: 12x WD Red 4TB NAS

 

Any idea what's wrong?

 

Mod edit: change icon

Syslog.txt

Link to comment

Moving from v5 to v6 means having to move to 64 bit drivers, instead of the 32 bit ones we've been using.  We've had surprisingly little trouble, considering, so far, but it looks like you have had the misfortune of requiring a driver (aacraid for your Adaptec card) who's 64 bit version is not working correctly.  It appears to be hanging rather badly.  I don't expect -beta8 to have a newer version either.  Tom will need to find a newer version of aacraid for you.  Until then, I don't think you are going to be able to use the v6 beta releases.  Sorry.  :(

 

Others have had problems with missing disks because of the Marvell drivers, but that has been resolved in a recent beta release.

Link to comment

It's not a widespread problem, but there have been several posts recently about "missing" disks that are apparently due to this issue (driver issues with certain cards).

 

Tom is aware of these, and I'm sure this will be resolved over time ... I'd just use the v5 stable for now, and check with each Beta release to see if it's been resolved for your specific card.    The Release Notes always show what's changed ... so you should be able to tell without the need to actually do the upgrade.

 

Link to comment

... The Release Notes always show what's changed ... so you should be able to tell without the need to actually do the upgrade.

 

While the Release Notes do show what UnRAID features are changed and fixes added and new drivers Tom has added, and a significantly different kernel used, but they don't mention all of the changes in the kernel if that was changed.  That is up to the kernel guys, Linus Torvalds et al (somewhere online there are Release Notes for each kernel release).  Most of the driver modules included in an UnRAID release are simply those included with the specified kernel release.  So I'm afraid you probably won't know if a particular driver is updated without testing the UnRAID release yourself.

Link to comment

Hey RobJ,

 

I'm having a similar issue with my Areca Card, clearly its a driver related issue but I'm unsure how to resolve this, I know the drivers are available for download via the areca website but no idea how to rectify this myself:

 

http://www.areca.com.tw/support/s_linux/linux.htm

 

From your other thread, I think you're doing what you can.  That site shows what the version should be, but we need to know what is currently included, and what previously worked for you.

 

I'm going to post some guidelines shortly, for those in similar situations.

Link to comment

Guidelines when your driver used to work, but doesn't any more.

 

1. Check for a firmware upgrade for your card, or BIOS update if chipset is on motherboard.  It's possible that changes in more recent kernels have caused new issues with an older firmware.  If you are going from 32 bit to 64 bit, the old firmware may not be compatible, or may not have been well tested in a 64 bit system.  By now, issues with your card may have been identified and fixed.

 

2. Try disabling VT-d or IOMMU for your motherboard, and testing again.  There are at least two reports of users successfully accessing all cards and drives once VT-d or IOMMU was disabled for their motherboard.

 

3. Provide a syslog for your system using the very latest UnRAID release (at this writing, v6.0-beta8).  We need to see what the current UnRAID software and kernel are seeing of your hardware, and what versions of the driver modules are loading.  Zip the syslog if necessary.

 

4. Provide a syslog from the very last version of UnRAID where your card ran well.  We need to see what the kernel and drive saw then, and what the version of the driver was, that ran successfully.  Zip the syslog if necessary.

 

5. Provide the exact model name of the card/chipset, and if possible what version of firmware is loaded.

 

6. Briefly describe what is failing, "drives missing", "card not recognized", etc.

 

These may help us to help you, and more importantly may help Tom determine what needs to be done.

 

Please understand that if a firmware update doesn't help, then there is probably nothing that anyone can do to help you run the CURRENT UnRAID version.  You will have to wait for a FUTURE version, with better driver support for your card.  Please also understand that Tom may not be able to help in any way, as it is generally up to the vendors or other outside groups to provide these drivers.

Link to comment

So I finally got around to updating the fw on the adaptec card and now all the disk are showing up and I can assign them. BUT there is a new problem... My previous fw did not list the model/serial number on the disks correctly they only had a number like this: 350014ee20aacd511 this was my parity disk btw. Now with the new fw the disk show up with model and a serial number like this: WDC_WD40EFRX-68WT0N0-WD-WCC4E6C7VTYK

 

Now the problem is that non of the serial numbers match the previous setup so therefor all the disk come up as wrong when I try to assign them. This of course happens when I go back to 5.0.5 because of the new fw.. I have attached a picture so you can see. From the picture you can see the new and old serial numbers . Is there any way to fix this besides just make a hole new array and use my backups? Thank GOD I have the backup.. :) Or just go back to the previous fw and stay on 5.0.5 forever..?

06-09-2014_18-55-20.jpg.1de246bca5ed71ecc044e823b5317ee8.jpg

Link to comment

I would suggest the following:

  • Make sure you know which is the serial number for parity disk
  • Do a 'new config' via the options on the Utility tab
  • Assign all the drives EXCEPT the parity drive
  • Start the array and check all the drives look OK.  If any drive shows as 'unformatted' that almost certainly means that you made a mistake and that is your parity drive.
  • If it all looks OK, you can then stop the array; assign the parity disk; and let new parity be calculated

Link to comment

Ok, I'll try that. I've never used the new config option so not entirely sure what it does, but I'll read up on that before I do it. Either way this issue seems to be resolved now that i can assign all my disks, and I'll get the array up an running soon when I've sorted out this new issue. So I'll just rename this thread to solved, boot up beta8 and check if everything else works. Thank you all for your help! :)

Link to comment

I would suggest the following:

  • Make sure you know which is the serial number for parity disk
  • Do a 'new config' via the options on the Utility tab
  • Assign all the drives EXCEPT the parity drive
  • Start the array and check all the drives look OK.  If any drive shows as 'unformatted' that almost certainly means that you made a mistake and that is your parity drive.
  • If it all looks OK, you can then stop the array; assign the parity disk; and let new parity be calculated

 

Just curious, would it work instead if you simply unassigned the parity drive, then started and stopped the array, then re-assigned the parity drive and started the array, letting it build parity?  I haven't tried it, and I confess I don't want to try it, but does someone know if that would work?  You would not have to use 'New config' and it would save having to record and re-assign all of the data drives, and would ensure that the parity drive is correctly assigned.  If after unassigning the parity drive it still shows the rest of the drives as wrong, then it won't work, and you can forget I suggested this.  It's just that without a parity drive, it doesn't matter if drives are wrong, so shouldn't mark them as wrong.

Link to comment

So I finally got around to updating the fw on the adaptec card and now all the disk are showing up and I can assign them. BUT there is a new problem... My previous fw did not list the model/serial number on the disks correctly ...

 

That's actually very good news.  Yes, there's a minor issue because the new firmware is working correctly and the old one didn't, but once you have the array 'fixed', you are better off.

 

Can you list the exact model and firmware here, for others who may have the same issue.  Only a small number of users load the betas, so it would be good to have this info ready, once v6.0 goes final.

 

By the way, I've updated the Guidelines above, to try disabling VT-D.

Link to comment

I would suggest the following:

  • Make sure you know which is the serial number for parity disk
  • Do a 'new config' via the options on the Utility tab
  • Assign all the drives EXCEPT the parity drive
  • Start the array and check all the drives look OK.  If any drive shows as 'unformatted' that almost certainly means that you made a mistake and that is your parity drive.
  • If it all looks OK, you can then stop the array; assign the parity disk; and let new parity be calculated

 

Just curious, would it work instead if you simply unassigned the parity drive, then started and stopped the array, then re-assigned the parity drive and started the array, letting it build parity?  I haven't tried it, and I confess I don't want to try it, but does someone know if that would work?  You would not have to use 'New config' and it would save having to record and re-assign all of the data drives, and would ensure that the parity drive is correctly assigned.  If after unassigning the parity drive it still shows the rest of the drives as wrong, then it won't work, and you can forget I suggested this.  It's just that without a parity drive, it doesn't matter if drives are wrong, so shouldn't mark them as wrong.

No.  unRAID thinks all the disk serial numbers are wrong which is why you have to go via the new config route.

 

In theory when going that route you could also initially assign the parity and tick the box to say it is already valid.  This would save rebuilding parity.  However if you accidentally assigned one of the data disks as parity then on starting the array unRAID would immediately try and start building parity on it and destroy the existing data (from previous forum messages this has happened a number of times).    This is why I recommended starting without parity assigned - it gives you a chance to notice the incorrect assignment and rectify it before any damage is done.  I think the chance of a failure while rebuilding parity after checking the assignment is correct is lower than the chance of a human error while assigning the disks.

Link to comment

Yeah, like itimpi said it didn't work, but the new config worked like a charm and the array is now up an running with a Parity-Sync in progress! :)

 

So I finally got around to updating the fw on the adaptec card and now all the disk are showing up and I can assign them. BUT there is a new problem... My previous fw did not list the model/serial number on the disks correctly ...

 

That's actually very good news.  Yes, there's a minor issue because the new firmware is working correctly and the old one didn't, but once you have the array 'fixed', you are better off.

 

Can you list the exact model and firmware here, for others who may have the same issue.  Only a small number of users load the betas, so it would be good to have this info ready, once v6.0 goes final.

 

By the way, I've updated the Guidelines above, to try disabling VT-D.

 

The model is Adaptec RAID 71605E (I've updated the thread title so it includes this so it's easier to find for other users). I bought the card in july but the fw version on the card was actually from may 2013 (fw version: 7.2.0 Build 30260). I should also mention that there were 2 other problems with this old fw 1: The temperature on the disks weren't showing in the stock web ui, but it worked in unmenu. 2: The status indicator for the disks was always showing just green like it was spinning.

 

The new fw that is working correctly is v7.5.0 Build 32033 from aug 2014 and can be downloaded from: http://www.adaptec.com/en-us/downloads/bios_fw/bios_fw_ver/productid=sas-71605e&dn=adaptec+raid+71605e.html   

Link to comment
  • 1 year later...

Update on Adaptec 71605, I have model 71605Q bought used from a recycle store. 16 drives on 4 sas cables. Running on unRaid 6.1.6

 

Flash the 71605 controller card bios to version 32033. I found it easiest to create a bootable USB drive, then unzip the download file onto the USB drive. Assuming only 1 (ONE) sata controller you have installed is the one you want to flash, boot the PC to dos from the USB drive, then type: "afu update" (without the quotes). This will start the sata controller card bios update and then verify, then tell you when to reboot.

 

To date, my experimenting has shown only the 32033 bios to work properly. One last tip, when writing the afu.exe and the bios file to your USB drive, make certain you use the two files from the same downloaded zip file. I found you can't simply overwrite the bios bin file and run it. The AFU.exe must be hard-coded to update only the bin file it was packaged with.

 

Edit: After successful flash and reboot, access the controller bios by keying ctrl-a when the controller is booting. Edit the controller and change to "HBA" mode.

  • Like 1
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.