Ugh...I removed the 2 "bad" drives and now I can't boot-up. Oh crap, I probably needed to remove them from unRAID before physically removing them from the case?

 

uMP

 

OMG...just put the drives back in and still can't get to the array! This might be a dumb question, but do they have to go back to the exact same sata positions??

 

EDIT- Oops...too impatient; got the array back. So...remove the "bad" drives from unRAID and then physically remove from case?

 

EDIT2- OK...so maybe I do need to get the drives back on the same sata cables...because now the array is back up, but while I guess I got lucky with disk 2, it is now showing one of the "new" drives in the disk 5 position...and it shows the "bad" disk 5 as unassigned. What do I need to do; get the "bad" disk 5 back on the right sata cable, unassign both "bad" drives and remove them? Or can I do something easier than that?

 

uMP

If you took a screen shot from before you started of the devices page you can just go to the devices page and assign the correct drives to their respective slots.

 

Well, it's seeing the "new" .5 as disk 5 Joe, so I think I just need to switch it with the old, "bad" disk 5 and I should be OK. So...just for the record...how do I unassign them from unRAID before I physically remove them from the box and do the restore?

Doesn't the restore remove them on it's own?  So power down, remove bad drives, add new drives, powerup, go to devices, assign new drives, restore array..?

Doesn't the restore remove them on it's own?  So power down, remove bad drives, add new drives, powerup, go to devices, assign new drives, restore array..?

 

I don't think I could get the Web UI without the "bad" drives? Unless I just didn't give it enough time. So...should I remove them from the array or just physically remove them; or does it matter?

 

uMP

So...should I remove them from the array or just physically remove them; or does it matter?

 

I have no idea   Haven't had to do any of this yet!

 

JL is the man

Well...I just went and switched my drives. I'm going to check and see if they're in the "right" positions now. It may not have even been necessary, but I figured it couldn't hurt. I'll wait to hear from Joe on what comes next.

 

uMP

If you expect to re-calc parity, and all the data drives you care for are attached, and assigned, then you can press the button labeled as "restore"  Just don't press "restore" if you are expecting to re-construct a disk from parity...(pressing it invalidates parity)

 

It will immediately set a new disk configuration based on the assigned and working disks

(and immediately invalidate the parity calc based on the old configuration).

 

When you then start the array it will perform a full parity calc on the currently assigned drives.

 

As mentioned in prior posts, make sure you also do a parity check after you do the initial parity calc. (The initial calc writes to the disk, but the check reads it to be sure the disk can be read back accurately)

If you expect to re-calc parity, and all the data drives you care for are attached, and assigned, then you can press the button labeled as "restore"   Just don't press "restore" if you are expecting to re-construct a disk from parity...(pressing it invalidates parity)

 

It will immediately set a new disk configuration based on the assigned and working disks

(and immediately invalidate the parity calc based on the old configuration).

 

When you then start the array it will perform a full parity calc on the currently assigned drives.

 

As mentioned in prior posts, make sure you also do a parity check after you do the initial parity calc. (The initial calc writes to the disk, but the check reads it to be sure the disk can be read back accurately)

 

Yikes; sorry, in English please...lol. I have 2 "bad" drives I'm removing Joe. I have copied their data onto a new drive that hasn't been assigned in the array yet. You guys said remove the "bad" drives and "restore" to keep all data, but with a rebuilt array.

 

So I'm trying to figure out if I need to remove the "bad" drives from the array, then physically remove them from the box...or just physically remove them and restore?

 

As it stands, they're in the box...but now I can't get my Web UI to come up. I may just need to wait it out.

 

EDIT- It's been a while, and still no Web UI.  :'( Man, this is confusing. First, with drives out, Web UI wouldn't come up. So I put them back in, but in the "wrong" positions...and it eventually came up. So I swapped the 2 that were wrong...now it won't come up again.  

 

EDIT2- OK, put a monitor back on the box...watched the boot; everything seemed to run OK, and now the Web UI is back up. Huh, needed a hard-boot for some reason. So I'm back to my original question: remove "bad" drives from array, remove physically, and restore...or just remove physically and restore?

 

uMP

If you expect to re-calc parity, and all the data drives you care for are attached, and assigned, then you can press the button labeled as "restore"   Just don't press "restore" if you are expecting to re-construct a disk from parity...(pressing it invalidates parity)

 

It will immediately set a new disk configuration based on the assigned and working disks

(and immediately invalidate the parity calc based on the old configuration).

 

When you then start the array it will perform a full parity calc on the currently assigned drives.

 

As mentioned in prior posts, make sure you also do a parity check after you do the initial parity calc. (The initial calc writes to the disk, but the check reads it to be sure the disk can be read back accurately)

 

Yikes; sorry, in English please...lol. I have 2 "bad" drives I'm removing Joe. I have copied their data onto a new drive that hasn't been assigned in the array yet. You guys said remove the "bad" drives and "restore" to keep all data, but with a rebuilt array.

 

So I'm trying to figure out if I need to remove the "bad" drives from the array, then physically remove them from the box...or just physically remove them and restore?

 

As it stands, they're in the box...but now I can't get my Web UI to come up. I may just need to wait it out.

 

EDIT- It's been a while, and still no Web UI.  :'( Man, this is confusing. First, with drives out, Web UI wouldn't come up. So I put them back in, but in the "wrong" positions...and it eventually came up. So I swapped the 2 that were wrong...now it won't come up again.  

 

EDIT2- OK, put a monitor back on the box...watched the boot; everything seemed to run OK, and now the Web UI is back up. Huh, needed a hard-boot for some reason. So I'm back to my original question: remove "bad" drives from array, remove physically, and restore...or just remove physically and restore?

 

uMP

If all of your data has been copied to good drives, you can un-assign the bad drives, then power down, then remove them.

When you power up the array will complain they are missing.  Use the devices page to assign the new drive(s) (if they are not yet assigned)

then press restore to Set a New Disk Configuration and begin the process of initially calculating parity on the new disk configuration.

 

(In reality, unless the bad drives were using ports on disk controllers you need, or causing interference on the SATA bus, or loading down the power supply, you can leave the task of physically removing the bad drives until the very end, after they have been un-assigned and parity calc'd with the new.  It is up to you.  As long as they are logically un-assigned, and the replacement drive(s) assigned you can set a new disk configuration and calculate parity based on the new configuration)

Yeah, I might as well face it...my data is all F'd up. I just went to stop the array...wouldn't refresh, Web UI just hung; now I can't get the Web UI back. Probably another hard-boot, and that'll be that; my hope of getting data back is nill.

 

uMP

Yeah, I might as well face it...my data is all F'd up. I just went to stop the array...wouldn't refresh, Web UI just hung; now I can't get the Web UI back. Probably another hard-boot, and that'll be that; my hope of getting data back is nill.

 

uMP

 

OK, with a monitor back on the box, kept getting errors on sata3...misconfigured, link is online but reset failed; something like that. It took forever, but finally got to root sign-in...but the Web UI is mostly disabled (can't get to lots of links)...and it doesn't look like unRAID sees ANY of my hard drives.

 

Go ahead, I'm screwed right?

 

Mar 21 13:02:06 unSERVER kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Mar 21 13:02:06 unSERVER kernel: ata3.00: link online but device misclassifed

Mar 21 13:02:06 unSERVER kernel: ata3: link online but 1 devices misclassified, retrying

Mar 21 13:02:06 unSERVER kernel: ata3: reset failed (errno=-11), retrying in 5 secs

Mar 21 13:02:06 unSERVER kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Mar 21 13:02:06 unSERVER kernel: ata3.00: link online but device misclassifed

Mar 21 13:02:06 unSERVER kernel: ata3: link online but 1 devices misclassified, retrying

Mar 21 13:02:06 unSERVER kernel: ata3: reset failed (errno=-11), retrying in 5 secs

Mar 21 13:02:06 unSERVER kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Mar 21 13:02:06 unSERVER kernel: ata3.00: link online but device misclassifed

Mar 21 13:02:06 unSERVER kernel: ata3: link online but 1 devices misclassified, retrying

Mar 21 13:02:06 unSERVER kernel: ata3: reset failed (errno=-11), retrying in 30 secs

Mar 21 13:02:06 unSERVER kernel: ata3: limiting SATA link speed to 1.5 Gbps

Mar 21 13:02:06 unSERVER kernel: atiixp 0000:00:14.1: IDE controller (0x1002:0x439c rev 0x00)

Mar 21 13:02:06 unSERVER kernel: ATIIXP_IDE 0000:00:14.1: PCI INT A -> GSI 16 (level, low) -> IRQ 16

Mar 21 13:02:06 unSERVER kernel: atiixp 0000:00:14.1: not 100%% native mode: will probe irqs later

Mar 21 13:02:06 unSERVER kernel: ide0: BM-DMA at 0xfa00-0xfa07

Mar 21 13:02:06 unSERVER kernel: atiixp 0000:00:14.1: simplex device: DMA disabled

Mar 21 13:02:06 unSERVER kernel: ide1: DMA disabled

 

uMP

Go ahead, I'm screwed right?

 

Remember the nice thing about unraid is each disk has it's own file system.  So if worst came to worst, you can get your files off each disk via your windows computer, unlike my crappy qnap fiasco where all the data was striped.  But I'm sure JL can help you figure it out without going to that extreme.  Totally just curious, but was the mobo you have one of the supported ones?  Sounds like you're running into a lot of problems

Go ahead, I'm screwed right?

 

Remember the nice thing about unraid is each disk has it's own file system.  So if worst came to worst, you can get your files off each disk via your windows computer, unlike my crappy qnap fiasco where all the data was striped.  But I'm sure JL can help you figure it out without going to that extreme.  Totally just curious, but was the mobo you have one of the supported ones?  Sounds like you're running into a lot of problems

 

I think it was too new to officially be on the list Shawn, but it was recommended, and I think GigaByte's, in general, play plenty nice with unRAID.

 

I just think that one drive has finally bit the dust. It was perfectly fine for what, a month or so.

 

uMP

Yeah, I might as well face it...my data is all F'd up. I just went to stop the array...wouldn't refresh, Web UI just hung; now I can't get the Web UI back. Probably another hard-boot, and that'll be that; my hope of getting data back is nill.

 

uMP

 

OK, with a monitor back on the box, kept getting errors on sata3...misconfigured, link is online but reset failed; something like that. It took forever, but finally got to root sign-in...but the Web UI is mostly disabled (can't get to lots of links)...and it doesn't look like unRAID sees ANY of my hard drives.

 

Go ahead, I'm screwed right?

 

Mar 21 13:02:06 unSERVER kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Mar 21 13:02:06 unSERVER kernel: ata3.00: link online but device misclassifed

Mar 21 13:02:06 unSERVER kernel: ata3: link online but 1 devices misclassified, retrying

Mar 21 13:02:06 unSERVER kernel: ata3: reset failed (errno=-11), retrying in 5 secs

Mar 21 13:02:06 unSERVER kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Mar 21 13:02:06 unSERVER kernel: ata3.00: link online but device misclassifed

Mar 21 13:02:06 unSERVER kernel: ata3: link online but 1 devices misclassified, retrying

Mar 21 13:02:06 unSERVER kernel: ata3: reset failed (errno=-11), retrying in 5 secs

Mar 21 13:02:06 unSERVER kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Mar 21 13:02:06 unSERVER kernel: ata3.00: link online but device misclassifed

Mar 21 13:02:06 unSERVER kernel: ata3: link online but 1 devices misclassified, retrying

Mar 21 13:02:06 unSERVER kernel: ata3: reset failed (errno=-11), retrying in 30 secs

Mar 21 13:02:06 unSERVER kernel: ata3: limiting SATA link speed to 1.5 Gbps

Mar 21 13:02:06 unSERVER kernel: atiixp 0000:00:14.1: IDE controller (0x1002:0x439c rev 0x00)

Mar 21 13:02:06 unSERVER kernel: ATIIXP_IDE 0000:00:14.1: PCI INT A -> GSI 16 (level, low) -> IRQ 16

Mar 21 13:02:06 unSERVER kernel: atiixp 0000:00:14.1: not 100%% native mode: will probe irqs later

Mar 21 13:02:06 unSERVER kernel: ide0: BM-DMA at 0xfa00-0xfa07

Mar 21 13:02:06 unSERVER kernel: atiixp 0000:00:14.1: simplex device: DMA disabled

Mar 21 13:02:06 unSERVER kernel: ide1: DMA disabled

 

uMP

The messages you show are all from 1 disk  (Affiliated with ata3).  Post a syslog so we can see the full picture.  Your data is likely fine.  You may have a bad motherboard, or a bad disk controller, or a cable not completely plugged in,  but the data itself is probably just fine.

 

 

OK...I must have knocked a sata cable loose or something. Opened the case up, double-checked everything; fired-up much quicker...and no yellow light on the HDD led.

 

Was able to get into Web UI, unassign the 2 "bad" drives, and shut down. Now I take them out, fire-up, assign 2 new drives, and do a "restore", right? You're sure it's a restore? And the data copied off the old, "bad" drives...onto the new 1.5...will be rebuilt and intact?

 

uMP

OK...I must have knocked a sata cable loose or something. Opened the case up, double-checked everything; fired-up much quicker...and no yellow light on the HDD led.

 

Was able to get into Web UI, unassign the 2 "bad" drives, and shut down. Now I take them out, fire-up, assign 2 new drives, and do a "restore", right? You're sure it's a restore? And the data copied off the old, "bad" drives...onto the new 1.5...will be rebuilt and intact?

 

uMP

Yes, this time you want to set a new disk configuration and calculate new parity based on the new configuration.  You are not expecting to use parity and the old data disks to reconstruct onto a replacement.  (you already did that)  So.. it is ok for you to press "restore"

OK...I must have knocked a sata cable loose or something. Opened the case up, double-checked everything; fired-up much quicker...and no yellow light on the HDD led.

 

Was able to get into Web UI, unassign the 2 "bad" drives, and shut down. Now I take them out, fire-up, assign 2 new drives, and do a "restore", right? You're sure it's a restore? And the data copied off the old, "bad" drives...onto the new 1.5...will be rebuilt and intact?

 

uMP

Yes, this time you want to set a new disk configuration and calculate new parity based on the new configuration.  You are not expecting to use parity and the old data disks to reconstruct onto a replacement.  (you already did that)  So.. it is ok for you to press "restore"

 

OK...old, "bad" disks out...assigned new disks, and restoring.

 

EDIT- The restore...the parity-sync...is expected to take about 8 hours on a 1.5T parity drive?

 

uMP

EDIT- The restore...the parity-sync...is expected to take about 8 hours on a 1.5T parity drive?

 

Sounds about right.

OK...I must have knocked a sata cable loose or something. Opened the case up, double-checked everything; fired-up much quicker...and no yellow light on the HDD led.

 

Was able to get into Web UI, unassign the 2 "bad" drives, and shut down. Now I take them out, fire-up, assign 2 new drives, and do a "restore", right? You're sure it's a restore? And the data copied off the old, "bad" drives...onto the new 1.5...will be rebuilt and intact?

 

uMP

Yes, this time you want to set a new disk configuration and calculate new parity based on the new configuration.  You are not expecting to use parity and the old data disks to reconstruct onto a replacement.  (you already did that)  So.. it is ok for you to press "restore"

 

OK...old, "bad" disks out...assigned new disks, and restoring.

 

EDIT- The restore...the parity-sync...is expected to take about 8 hours on a 1.5T parity drive?

 

uMP

Most people report read top speeds between 75 and 100  MB/s on a drive, slowing to 60 MB/s or less on inner cylinders.

Let's assume best case and it will be 1GB per 10 seconds, at 60MB/s it would be 16.6 seconds per 1Gig.  At 50 MB/s it will take 20 seconds per 1Gig.  Let's assume your motherboard has absolutely NO bottlenecks on its internal busses.

 

Now, writing to a disk is typically much slower than reading.  But for this exercise let's assume it can also be written to at 1GB per 10 seconds. (100 MB/s)

 

You have 1500GB (1.5TB) to write, therefore, 15000 seconds at 100MB/s...  = 4.16 hours

 

Now, if it cannot perform a sustained write at 100 MB/s...

                                                    It will take 24900 seconds at 60 MB/s...    = 6.91 hours

                                                              Or  30000 seconds at 50 MB/s...    = 8.32 hours

 

Remember, it will be slower than the time it takes to perform a full parity check.  The rebuild will be limited by the speed of the slowest drive involved and/or the capacity of the system bus to read all the data.  Don't forget you are reading 1.5TB each from each of your other disks.

 

8 hours does not sound unreasonable to me.  don't forget, I did not add any overhead to actually calculate parity.  The actual process will take less time if you have a mix of sizes of disks, as there will be less to read once you get past the size of the smaller disks.

 

Joe L.

OK, the parity-sync is done; what's next? All my data drives...an existing one on disk 1, and the two new ones which are at positions 2 and 3...say they are unformatted. Formatting sounds scary. What do I do to keep data intact?

 

uMP

OK, the parity-sync is done; what's next? All my data drives...an existing one on disk 1, and the two new ones which are at positions 2 and 3...say they are unformatted. Formatting sounds scary. What do I do to keep data intact?

 

uMP

Do NOT press the format button.  This is the bug everybody has been complaining about on 4.5.3.  The disks are being mis-reported.

(hopefully)

 

Stop the array and reboot, they should show up properly.

OK, the parity-sync is done; what's next? All my data drives...an existing one on disk 1, and the two new ones which are at positions 2 and 3...say they are unformatted. Formatting sounds scary. What do I do to keep data intact?

 

uMP

Do NOT press the format button.  This is the bug everybody has been complaining about on 4.5.3.  The disks are being mis-reported.

(hopefully)

 

Stop the array and reboot, they should show up properly.

 

Right On; wasn't about to press without checking first. Will stop array and reboot tonight. Keeping fingers crossed.

 

uMP

Just in case they show up unformatted again after a reboot, just stop the array and start it again (don't reboot).  My experience with the bug is that every time I reboot, all my disks show up as unformatted.  No amount of reboots will change that, only stopping and then starting the array.

Had a weird issue; I'm sure it's not that big a deal...just starting to notice some things that never seemed to happen before. When I got home from work, I tried to stop the array so I could reboot; the Web UI just hung on me...and I had to do a hard boot.   Head-scratcher.

 

Anyway, everything is back up; so, on first boot, it'll do a parity-check? How long will that take? (EDIT- actually, unMENU shows the status...2% done, 322 minutes to go) Man, I've been without my array forever. Also, one drive...a new .5, that I didn't set the file system on or copy anything to, is still legitimately unformatted. How do I handle that?

 

uMP

on first boot, it'll do a parity-check? How long will that take? (EDIT- actually, unMENU shows the status...2% done, 322 minutes to go) Man, I've been without my array forever. Also, one drive...a new .5, that I didn't set the file system on or copy anything to, is still legitimately unformatted. How do I handle that?

 

uMP

If the 500 Gig drive shows as unformatted, and it is the ONLY drive showing as unformatted, then it is safe to press the "Format" button on the web-interface.

unRAID will automatically run a parity check after a hard boot.  This is normal.  You may also see 2 parity errors, this is also normal.  If you see a lot of parity errors, then there may be a problem.