I thought I was ready to add my first drive to my new unRAID server, and I thought I'd just start at the top of the list of the Devices page. So I added the parity drive and the page started to reload. I waited for a minute or two, and then clicked on the main page, but my browser finally decided I couldn't access the page anymore. I can still telnet to the server, but I'm not sure what to check to see what's going on?

 

Can anyone offer any advice?

I thought I was ready to add my first drive to my new unRAID server, and I thought I'd just start at the top of the list of the Devices page. So I added the parity drive and the page started to reload. I waited for a minute or two, and then clicked on the main page, but my browser finally decided I couldn't access the page anymore. I can still telnet to the server, but I'm not sure what to check to see what's going on?

 

Can anyone offer any advice?

No idea what it might be doing.  You might read the sticky post at the top of this forum suggesting a syslog be attached when requesting help.  (It might give a clue what is happening)  It might be calculating parity from all the (non-existent) data drives.

Sorry, for being "that guy" that didn't follow the directions

 

Here's my config:

 

 

Motherboard -  ECS A885GM-A2 (V1.1)

Processor - AMD Sempron 145 Sargas 2.8GHz

Parity Drive - HITACHI Deskstar 5K3000

Hard Drives -  I have various other drives including Western Digital (EARS, EADS, and an EACS) Green Drives and some Samsung drives

RAM - 2 x Crucial 2GB 240-Pin DDR3 SDRAM 1333

PSU - CORSAIR CMPSU-850TX 850W (I believe this is a single rail supply)

Expansion Cards - SUPERMICRO AOC-SASLP-MV8

Flash Drive - Lexar JumpDrive FireFly 16GB USB 2.0 Flash Drive  (Probably should have ordered a spare of these)

EDIT:

unRAID version 4.7

I set the Hitachi drive as the parity drive.

 

Syslog attached.

syslog-2012-03-29-2125.txt

I don't see anythng specific in the syslog.  since there is no array yet, just hit the reset button on the server and reboot.

 

You do have one issue, but it probably will not matter too much unless you tried to use that drive as your parity drive.

One of your drives has an HPA.

Mar 29 20:11:34 Hyperion kernel: ata4.00: HPA detected: current 1953523055, native 1953525168

Wow. I missed that. I looked at the log and didn't see the server doing anything so I just rebooted and the web interface started working again. I've had to do that a couple time already, so it's beginning to make me a little nervous about what's going on there.

 

As far as the HPA goes, my OLD motherboard was a Gigabyte that I replaced due to the HPA issue. I've pre cleared all the drives currently connected with the exception of the one disk that I plan to copy my data from. Is there a way for me to verify which disk has the HPA? Obviously is "should" be the drive with my data on it as I haven't pre cleared it yet, but I'd like to double check if possible.

 

Thanks for taking the time to look through my syslog!

Well. So much for that. After a forum search it looks like preclear does NOT remove HPA from disks. So I'll have to figure out how to find and remove HPA from each disk that I plan to add my server. Fortunately, I've only got 10 disks. Unfortunately, I've got 10 disks

hdparm -N /dev/sdX

when run on each of your disks will tell you which have a different logical than physical size.

 

Then,

hdparm -N pXXXXXXXXXXXXX /dev/sdX

will set the HPA (or remove it) in many cases.  (Some disk controllers will allow this command to work, others will not)

You need to use the correct size instead of XXXXXXXXXXXX, but that is printed by the hdparm command.

Don't forget the leading "p" to make it know you want it to be permanent.

 

Lastly, once you make it the correct full size, it will not longer be considered as pre-cleared, since the signature was for the old size, and the new extended blocks have not been cleared.  So... preclear it once more if you want it to be recognized as a pre-cleared drive.

 

Joe L.

I tried this on a disk that should not have an HPA (2 TB parity disk that is new) and this is the result I got.

 

max sectors  = 3907029168/14715056(18446744073321613488?), HPA setting seems invalid (buggy kernel device driver?)

 

I didn't know what to expect, but this certainly wasn't it. Is this result indicative of an HPA? All three of the drives give similar messages with different values.

 

EDIT: Looking through the forums some more, here are the results from all three drives in my array.

 

root@Hyperion:~# hdparm -N /dev/sdb

 

/dev/sdb:

max sectors  = 3907029168/14715056(18446744073321613488?), HPA setting seems invalid (buggy kernel device driver?)

root@Hyperion:~# hdparm -N /dev/sdc

 

/dev/sdc:

max sectors  = 2930275055/11041584(18446744072344861488?), HPA setting seems invalid (buggy kernel device driver?)

root@Hyperion:~# hdparm -N /dev/sdd

 

/dev/sdd:

max sectors  = 1953525168/7368112(1953525168?), HPA setting seems invalid (buggy kernel device driver?)

 

Do I understand correctly that drive sdc is the one with the HPA? I wish this error was a little more descriptive. It makes it sound as though each of the three drives has an HPA.

yes, I too think it is /dev/sdc.

 

Since the hdparm command cannot get the true size you'll have to resort to a third party tool to reset the HPA.

 

I doubt you'll have success with the hdparm command itself. 

 

Joe L.

looking through the syslog, I actually found a second disk that was showing an HPA error. I followed the commands you provided above for each disk, and I used the native values given in my syslog, and after rebooting I don't get any HPA errors. Should I take that to mean that I was lucky and it correctly removed the HPA from both disks?

 

Also, since one of the disks I added to my array had an HPA, and I let it calculate the initial parity, should I start over completely? that is, remove all the disks, preclear everything, and start over? Or can I just preclear the two affected disks, add them back to my array, recalculate parity, and start copying disks?

 

It's a little overwhelming the amount you have to learn to get this going. I'm still hoping things get smoother once this is all up and running 

 

Thanks again for all of the help with my issues so far!

looking through the syslog, I actually found a second disk that was showing an HPA error. I followed the commands you provided above for each disk, and I used the native values given in my syslog, and after rebooting I don't get any HPA errors. Should I take that to mean that I was lucky and it correctly removed the HPA from both disks?

Yes.

Also, since one of the disks I added to my array had an HPA, and I let it calculate the initial parity, should I start over completely? that is, remove all the disks, preclear everything, and start over? Or can I just preclear the two affected disks, add them back to my array, recalculate parity, and start copying disks?

Either will work.  I would think unRAID would see the changed size if the disk is already in the array.

 

I've been reading about this while my drives are preclearing, and I'm wondering if I should run

 

initconfig

 

before reassigning the drive to the array. If I don't have to wait on that 2TB drive to preclear again, I'd rather not. But I want to make sure I've taken all the precautions I should for my first setup.

I've been reading about this while my drives are preclearing, and I'm wondering if I should run

 

initconfig

 

before reassigning the drive to the array. If I don't have to wait on that 2TB drive to preclear again, I'd rather not. But I want to make sure I've taken all the precautions I should for my first setup.

That will work too.  It will force a new parity calculation as it immediately invalidates parity on the old disk configuration.  Just remember to answer "Yes" to its prompt (with a capital "Y" and lower case "es")

 

Joe L.

I've had this problem, i put it down to DNS since \\IP_Address_Of_Server worked fine but \\tower failed, both in web client and shares. Just my $0.02.  . A reboot like the OP said seemed to get it to register on the local network again.

I was ready to mark this as "[sOLVED]" again, but I added a data disk to my array and the web interface "locked up" again. Rebooting cleared it up, but I'm at a loss as to what's going on.

I was ready to mark this as "[sOLVED]" again, but I added a data disk to my array and the web interface "locked up" again. Rebooting cleared it up, but I'm at a loss as to what's going on.

Adding a new data disk to an array already protected by parity will result in exactly the behavior you describe IF the drive has not been pre-cleared and does not still have a valid pre-clear signature.

 

The array is off-line until unRAID clears the drive.  (typically, for a large drive, 6 hours or more)

 

The next time it locks up (as you describe it) can you log in on the system console?  Can you log in via telnet?

Can you "ping" the server from your window's PC?

 

Can you grab a syslog BEFORE you reboot.

 

Joe L.

The drive was freshly pre-cleared, and the array was offline when I attempted to add the drive. Once I assigned the drive, the page appeared to try to reload (i.e., the progress bar in the address bar moved to maybe 1/4 of the way across and stopped). After waiting a while, I clicked on the main page link, but nothing happened. After a few more minutes the browser gave up and just loaded the "Cannot Load Page" message.

 

I'll try to be a little more patient next time, and post a syslog. I'm getting a little impatient as I've been fiddling with this thing for two weeks and have yet to copy any data to it. After rebooting, the drive appeared to have been added to the array (which was still stopped). So I went ahead and upgraded my server key (rebooted), and added the other drive that I removed the HPA from with no problems (rebooted again). I then ran initconfig via a telnet session (rebooted again  ). All drives showed up as blue, so I started the array, formatted the two recently added drives, and I'm now waiting on the parity drive to build. With any luck, I'll not have anymore issues to necessitate posting a syslog.

 

I appreciate the help!