DNS-323, current Alt-F firmware, and 3TB drives

[ScottK]

Hi all,

I recently bought a new Seagate 3TB drive, model ST3000DM001:
Seagate Barracuda ST3000DM001 3TB 7200 RPM 64MB Cache SATA 6.0Gb/s 3.5" Internal Hard Drive

(Actually got it from buying the "external" USB model, and opening up the case)

When installing this into the DNS-323 running the current ALT-F version of the firmware, it sees the drive, and even is ready to format the drive for me.

But the initial starting of the format fails, with quite a weird error message.

I do NOT have the error message right now, I can/will post it tonight when I get home from work.

I did nothing to the drive, in terms of removing the exiting format/partitions on the drive before I attempted to use it in Alt-F.
(Thus, it is formatted with NTFS, with about 100 Megs of Seagate docs/utils that they ship on the drive initiall)

Has anyone gotten a 3TB drive to work on Alt-F?

If not, and there is some beta, or some way to play around to make it work, let me know.
This disk is a virgin disk, and is going in a 2nd (and unused) DNS-323, thus it is available for me to muck with, without any fear of losing any existing data.

Thanks for any/all help!
Scott

[Joao Cardoso]

You have to calm down, Scott :-)

Not only you posted your message twice, not waited to reach home to see what the error message was and post it, not had the time to search the forum ;-) 

[ScottK]

Haha, actually, for some reason, I got some weird error when I posted the first time, so I figured I would click <Post> again.
Only when I got the same weird error back again, did I think to check to see if it actually posted.  (It did!)

I did try to search for this problem, and could only seem to come across one post, which some guy connected up a 3TB drive, and was able to view the NTFS files, but he didn't go any further?

I will search more now...

Let me know if you would like me to post the error message when I get home.  :)

Thanks again!
Scott

[ScottK]

Here is the error, and procedure to get it.

Select the disks you want to format

Format	Bay	Device	Capacity	Model
	right	sda	3000.6 GB	ST3000DM001-9YN166

Whirl your magic wand...

I want my disk as:
	Don't touch my disks in any way!
	One big filesystem per disk, for easy management (standard)
	Merge all disks in one big filesystem, low data security (JBOD)
	Maximum performance and space with two or three disks (one an external USB disk), but low data security (raid0)
	Data security, duplicate everything on two disks (and use an external USB disk, if available, as a spare) (raid1)
	Data security and space, with two disks plus an external USB disk (raid5)
And I want the filesystems to be:
	older, stable and faster (ext2)
	fast cleaning time, improved reliability (ext3)
	recent, faster cleaning time, best reliability, low fragmentation, big files support (ext4)

When I click on "Abracadabra", I get:

Stopping all services and disks... done.

GPT partitioning disk sda.../dev/sda Could not create partition 3 from 1565565839 to 1565565838 Could not change partition 3's type code to 8300! Error encountered; not saving changes. failed:

Information: Moved requested sector from 1565565839 to 1565565840 in
order to align on 8-sector boundaries.


At this point, the only option I have is to go "Back".


Any thoughts or suggestions?
Thanks again!
Scott

[Joao Cardoso]

Can you please go to Disk->Partitioner, select the disk in question and post what you see there?

What are your linux  command line skills?

I don't have a 3TB disk myself, only two user reports saying that it works OK. But they didn't say if they used the command line or the Wizard or Partiioner web pages.

I can't fix it if I don't know that it is damaged, and I can't fix it if I don't know how it is damaged :-O

Thanks 

[ScottK]

No problem!
This is a virgin disk, on a 2nd (and unused) DNS-323, so it is a perfect set up for us to work/play on.

Oh, I write Linux drivers for a living, so I definitely can play at the command line. :)


Here is the screenshot:

Select the disk you want to partition

fdisk: device has more than 2^32 sectors, can't use all of them

Partition	Bay	Device	Capacity	Model	Partition Table

right

sda

3000.6 GB

ST3000DM001-9YN166

fdisk: device has more than 2^32 sectors, can't use all of them

Partition right disk, 3000.6 GB, ST3000DM001-9YN166

Using GPT partitioning.

Every disk must have a swap partition as its first partition, 0.5GB is generally enought.

Keep	Dev	Start sector	Lenght	Size (GB)	Type
	sda1
	sda2
	sda3
	sda4
			Free:

I didn't change/mod anything, so the above is what it just decided to display when going to that screen.

Thanks for your help here, and thanks for the firmware!
Scott

[Joao Cardoso]

So the disk wizard (the 'sgdisk' program, really) performed part of its duties, but badly.

It should:

1-create a 500MB swap partition, which it did, 

2-then a second linux partition with the remaining disk space, which it did only partially (this second partition does not uses all the remaining disk space)

3-Then, if there is still some disk free space it should create a third partition -- and the error comes from this step.

This is either:

-a bug in the sgdisk program,

-an issue with the disk way of reporting/using the 4k sectors it uses (Seagate says that it uses Seagate SmartAlign™ technology for no-hassle transition to Advanced Format 4K sector technology)

-or a problem compiling sgdisk for Alt-F...

I might have a solution for you, but first I would like to collect some information from the current disk state, if you don't mind.

Please telnet or ssh the box as the 'root' user, no quotes, using the same password as you use to access its web pages,

then execute the following commands and post its output (you can just copy/paste, hope there is no line breaks issues):

   fdisk -l /dev/sda

   sfdisk -luS /dev/sda

   sgdisk -p /dev/sda

   sgdisk --first-in-largest /dev/sda

   sgdisk --end-of-largest /dev/sda

# The next steps will reformat the disk again:

   sgdisk --zap-all /dev/sda # erases MBR and GPT partition info

   sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda # creates a first 512MB swap partition

   sgdisk -p /dev/sda # reports the current disk partition

   sgdisk --first-in-largest /dev/sda # reports the disk next free start sector

   sgdisk --end-of-largest /dev/sda # reports the end of the largest area

   sgdisk --set-alignment=8 --new=2:0:0 --typecode=2:8300 /dev/sda # creates a second linux part using all remaining disk

   sgdisk -p /dev/sda

As this point you should have a good partition table, but probably you will have an identical one to what the disk wizard did.

I hope that the collected information will be enough to diagnose the issue

If you still have a 800GB second partition, let's see the disk partitioner can fix it.

Using the Disk Partitioner, uncheck the 'Keep' checkbox in the second partition (sda2), and fill in '3000' (without quotes) in its 'Size' entry field.

Adjust the value until the 'Free' size is lightly positive, then hit the 'Partition' button -- did it works? Do you now have a 3TB partition?

If it worked, you can go to Disk Filesystem and create a ext4 filesystem on the sda2 partition ('New FS Operation', select ext4 in the 'New FS', then 'Create' under 'Operation').

But don't do it now, as I would like to know how to create a big partition in the wizard.

Again at the command line, (this will repartition your disk again):

   sgdisk --zap-all /dev/sda

   sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda

   sgdisk --set-alignment=8 --new=2:0:+2900M --typecode=2:8300 /dev/sda

   sgdisk -p /dev/sda 

Any error? Going to the Disk Partitioner now, does it now shows a (+/-) 3GB second linux partition?

Yes, I know, lots of commands, but with your linux device driver experience it will be like a piece of cake :-)

Hope the time lag between you and me does not affects the interaction,

Thanks,

Joao

[ScottK]

Hi Joao!
 
Today I brought the DNS-323 device into work.
So we can play all day if needed!  (Shhhh, don't tell my boss, haha!) :)


Below are the commands and output:

# fdisk -l /dev/sda

fdisk: device has more than 2^32 sectors, can't use all of them

Found valid GPT with protective MBR; using GPT

Disk /dev/sda: 4294967295 sectors, 4095M
Logical sector size: 512
Disk identifier (GUID): a3462a1d-6cb7-47f7-96bf-1265d4705ed9
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134

   Device  Start (sector)    End (sector)  Size (sectors)   Code  Name
/dev/sda1              64         1048639         1048576   8200 
/dev/sda2         1048640      1565565838      1564517199   8300 

# sfdisk -luS /dev/sda

Disk /dev/sda: 97451 cylinders, 255 heads, 63 sectors/track

Warning: The partition table looks like it was made
  for C/H/S=*/256/63 (instead of 97451/255/63).
For this listing I'll assume that geometry.
Units = sectors of 512 bytes, counting from 0

   Device Boot    Start       End   #sectors  Id  System
/dev/sda1             1         - 4294967295  ee  Intel EFI GUID Partition Table
/dev/sda2             0         -          0   0  Empty
/dev/sda3             0         -          0   0  Empty
/dev/sda4             0         -          0   0  Empty
#

# sgdisk -p /dev/sda
Disk /dev/sda: 5860533168 sectors, 2.7 TiB
Logical sector size: 512 bytes
Disk identifier (GUID): A3462A1D-6CB7-47F7-96BF-1265D4705ED9
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134
Partitions will be aligned on 64-sector boundaries
Total free space is 4294967326 sectors (2.0 TiB)

Number  Start (sector)    End (sector)  Size       Code  Name
   1              64         1048639   512.0 MiB   8200 
   2         1048640      1565565838   746.0 GiB   8300 

# sgdisk --first-in-largest /dev/sda
1565565839

# sgdisk --end-of-largest /dev/sda
5860533134



# sgdisk --zap-all /dev/sda # erases MBR and GPT partition info
GPT data structures destroyed! You may now partition the disk using fdisk or
other utilities.

# sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda
Creating new GPT entries.
The operation has completed successfully.

# sgdisk -p /dev/sda
Disk /dev/sda: 5860533168 sectors, 2.7 TiB
Logical sector size: 512 bytes
Disk identifier (GUID): E9D88D02-D82E-4E80-A11F-E4E56A94EA64
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134
Partitions will be aligned on 64-sector boundaries
Total free space is 5859484525 sectors (2.7 TiB)

Number  Start (sector)    End (sector)  Size       Code  Name
   1              64         1048639   512.0 MiB   8200 

# sgdisk --first-in-largest /dev/sda
1048640

# sgdisk --end-of-largest /dev/sda
5860533134

# sgdisk --set-alignment=8 --new=2:0:0 --typecode=2:8300 /dev/sda
The operation has completed successfully.

# sgdisk -p /dev/sda
Disk /dev/sda: 5860533168 sectors, 2.7 TiB
Logical sector size: 512 bytes
Disk identifier (GUID): E9D88D02-D82E-4E80-A11F-E4E56A94EA64
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134
Partitions will be aligned on 64-sector boundaries
Total free space is 4294967326 sectors (2.0 TiB)

Number  Start (sector)    End (sector)  Size       Code  Name
   1              64         1048639   512.0 MiB   8200 
   2         1048640      1565565838   746.0 GiB   8300 



Scott: Yup, at this point, the partitions look exactly the same as before.


Next, I went in and did your next set of steps.
Go to the GUI -> Dis Partitioner, untick the Keep for sda2, and then adjust size to 3000.
(The remaining size was .056, so perfect!)

Click on "Partition" button.

It paused for a few seconds, and gave me:

fdisk: device has more than 2^32 sectors, can't use all of them
HTTP/1.0 200 OK
Content-Type: text/html; charset=UTF-8

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html style="height: 100%;"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title></title></head><body style="height: 100%;">
<center><h2>Disk Partitioner</h2></center>
<script type="text/javascript">
document.body.style.cursor = 'wait';
</script>
<p>Stopping disk...
 done.</p>
<p>Partitioning disk sda...
 done</p><p>Setting up partitions details...
 done</p><p>Reloading all disks...
 done</p>
<script type="text/javascript">
document.body.style.cursor = '';
</script>
<p><strong>Success</strong></p>
<script type="text/javascript">
function err() {
window.location.assign(document.referrer)
}
setTimeout("err()", 3000);
</script>
</body></html>

So the GUI got a little messed up, BUT!!!
IT WORKED!!!! :)

# sgdisk -p /dev/sda
Disk /dev/sda: 5860533168 sectors, 2.7 TiB
Logical sector size: 512 bytes
Disk identifier (GUID): E9D88D02-D82E-4E80-A11F-E4E56A94EA64
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134
Partitions will be aligned on 64-sector boundaries
Total free space is 109525 sectors (53.5 MiB)

Number  Start (sector)    End (sector)  Size       Code  Name
   1              64         1048639   512.0 MiB   8200 
   2         1048640      5860423639   2.7 TiB     8300 


The UI, when going back to Partitioner, also shows that it worked:

fdisk: device has more than 2^32 sectors, can't use all of them

Partition right disk, 3000.6 GB, ST3000DM001-9YN166

Using GPT partitioning.

Every disk must have a swap partition as its first partition, 0.5GB is generally enought.

Keep	Dev	Start sector	Lenght	Size (GB)	Type
	sda1
	sda2
	sda3
	sda4

So I think we are good here! :)

Is there any other data points you would like to grab from me?
Let me know, like I said, this disk and DNS-323 is completely a spare right now, so I am happy to do whatever you need to collect more information!

Thanks again for all your help!
Scott

[ScottK]

Added commentary, comments:

Since we are doing this "behind the scenes", the disk isn't actually being formatted.
(Just partitioned...)

So I went into the "Filesystem Maintenance", and attempt to format them.
The first one, I believe is mentioned to be swap, I left that alone.
The 2nd one, I am attempting to format as ext4, and when I do, I get an error from "awk".


Here it is before:

sda1	0.5GB	none
sda2	2794.0GB	none

When selecting "ext4" and "format", I get this:

awk: cmd. line:1: Division by zero

Dev.	Size	FS	Mnt	Label	Mount Options	FS Operations	New FS Operations
sda1	0.5GB	none
sda2	2794.0GB	formating...

However, I don't believe it is actually formatting...

Is there a command line you would like me to try with this one?

Thanks!
Scott

[Joao Cardoso (Alt-F)]

On Wednesday 02 May 2012 08:06:58 ScottK wrote:
> On Wednesday, May 2, 2012 9:48:59 AM UTC-5, Joao Cardoso wrote:

...

> Below are the commands and output:

...

> # sgdisk --set-alignment=8 --new=2:0:0 --typecode=2:8300 /dev/sda
> The operation has completed successfully.

It has not, as the next command output shows.
The '--new=2:0:0' in the command above means "create the second partition
with all remaining disk space", which it did not.

As the sgdisk manual says: "A start or end value of 0 specifies the default
value, which is the start of the largest available block for the start sector
and the end of the same block for the end sector."

So it is a sgdisk bug, and I will report it back to its author.

> # sgdisk -p /dev/sda
> Disk /dev/sda: 5860533168 sectors, 2.7 TiB
> Logical sector size: 512 bytes
> Disk identifier (GUID): E9D88D02-D82E-4E80-A11F-E4E56A94EA64
> Partition table holds up to 128 entries
> First usable sector is 34, last usable sector is 5860533134
> Partitions will be aligned on 64-sector boundaries
> Total free space is 4294967326 sectors (2.0 TiB)
>
> Number Start (sector) End (sector) Size Code Name
> 1 64 1048639 512.0 MiB 8200
> 2 1048640 1565565838 746.0 GiB 8300
>
>
>
> Scott: Yup, at this point, the partitions look exactly the same as before.

Yes, as expected.

> Next, I went in and did your next set of steps.
> Go to the GUI -> Dis Partitioner, untick the Keep for sda2, and then adjust
> size to 3000.
> (The remaining size was .056, so perfect!)
>
> Click on "Partition" button.
>
> It paused for a few seconds, and gave me:
>
> fdisk: device has more than 2^32 sectors, can't use all of them
> HTTP/1.0 200 OK
> Content-Type: text/html; charset=UTF-8

...

> So the GUI got a little messed up,

Yes, I had already noticed it in the Partitioner screen dump, it is easy to
fix, its because the "fdisk: device has more than 2^32 sectors,..." message

> BUT!!! IT WORKED!!!! :)
>
> # sgdisk -p /dev/sda
> Disk /dev/sda: 5860533168 sectors, 2.7 TiB
> Logical sector size: 512 bytes
> Disk identifier (GUID): E9D88D02-D82E-4E80-A11F-E4E56A94EA64
> Partition table holds up to 128 entries
> First usable sector is 34, last usable sector is 5860533134
> Partitions will be aligned on 64-sector boundaries
> Total free space is 109525 sectors (53.5 MiB)
>
> Number Start (sector) End (sector) Size Code Name
> 1 64 1048639 512.0 MiB 8200
> 2 1048640 5860423639 2.7 TiB 8300

> So I think we are good here! :)

Yes!

> Is there any other data points you would like to grab from me?

Yes, please. To circunvent the sgdisk bug, I would like to see how to use the
command line to create the second partition. Please (this will again
repartition the disk, so afterwards you might have to re-adjust it in the
Partitioner, as you did before):

sgdisk --zap-all /dev/sda
sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda

end=$(sgdisk --set-alignment=8 --end-of-largest /dev/sda)
sgdisk --set-alignment=8 --new=2:0:${end} --typecode=2:8300 /dev/sda
sgdisk -p /dev/sda

Does it now shows a (+/-) 3TB second linux partition?
Could you please also use:

sgdisk --zap-all /dev/sda
sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda

start=$(sgdisk --set-alignment=8 --first-in-largest /dev/sda)
end=$(sgdisk --set-alignment=8 --end-of-largest /dev/sda)
sgdisk --set-alignment=8 --new=2:${start}:${end} --typecode=2:8300 /dev/sda
sgdisk -p /dev/sda

Also worked?

> Let me know, like I said, this disk and DNS-323 is completely a spare right
> now, so I am happy to do whatever you need to collect more information!
>
> Thanks again for all your help!

And thanks you for your availability!
Joao

> Scott

[Joao Cardoso]

Added commentary, comments:

Since we are doing this "behind the scenes", the disk isn't actually being formatted.
(Just partitioned...)

So I went into the "Filesystem Maintenance", and attempt to format them.
The first one, I believe is mentioned to be swap, I left that alone.
The 2nd one, I am attempting to format as ext4, and when I do, I get an error from "awk".

...

 

When selecting "ext4" and "format", I get this:

awk: cmd. line:1: Division by zero

Dev.	Size	FS	Mnt	Label	Mount Options	FS Operations	New FS Operations
sda1	0.5GB	none
sda2	2794.0GB	formating...

However, I don't believe it is actually formatting...

I think it is, the awk error is just Alt-F trying to compute the current elapsed percentage.

And you don't need to wait for completion on that page, you can go, e.g. to the Status page, then return to the Filesystem page...

As for the first (swap) partition, you have to use the command:

   mkswap /dev/sda1

   swapon /dev/sda1

ah, and you might be right, without a swap partition the creation of the filesystem might fail. If in the status page you don't see the completion percentage to increase, it is not working and you will have to repeat the procedure after creating the swap as above.

[ScottK]

 
No problem!


Here it the trace:
(It works!)

#  sgdisk --zap-all /dev/sda

GPT data structures destroyed! You may now partition the disk using fdisk or
other utilities.

# sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda
Creating new GPT entries.

The operation has completed successfully.

# end=$(sgdisk --set-alignment=8 --end-of-largest /dev/sda)

# echo $end
5860533134

#  sgdisk --set-alignment=8 --new=2:0:${end} --typecode=2:8300 /dev/sda

The operation has completed successfully.

# sgdisk -p /dev/sda
Disk /dev/sda: 5860533168 sectors, 2.7 TiB
Logical sector size: 512 bytes

Disk identifier (GUID): FDAB11D2-AE9D-4E03-9133-7D2C22B19879

Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134
Partitions will be aligned on 64-sector boundaries

Total free space is 30 sectors (15.0 KiB)

Number  Start (sector)    End (sector)  Size       Code  Name
   1              64         1048639   512.0 MiB   8200 

   2         1048640      5860533134   2.7 TiB     8300 




# sgdisk --zap-all /dev/sda

GPT data structures destroyed! You may now partition the disk using fdisk or
other utilities.

# sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda
Creating new GPT entries.

The operation has completed successfully.

# start=$(sgdisk --set-alignment=8 --first-in-largest /dev/sda)

# echo $start
1048640

#  end=$(sgdisk --set-alignment=8 --end-of-largest /dev/sda)

# echo $end
5860533134

# sgdisk --set-alignment=8 --new=2:${start}:${end} --typecode=2:8300 /dev/sda

The operation has completed successfully.

#  sgdisk -p /dev/sda
Disk /dev/sda: 5860533168 sectors, 2.7 TiB
Logical sector size: 512 bytes

Disk identifier (GUID): FFEB880C-BBB5-41DD-9BA0-45ED62F901AD

Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134
Partitions will be aligned on 64-sector boundaries

Total free space is 30 sectors (15.0 KiB)

Number  Start (sector)    End (sector)  Size       Code  Name
   1              64         1048639   512.0 MiB   8200 

   2         1048640      5860533134   2.7 TiB     8300 


Scott

[ScottK]

On Wednesday, May 2, 2012 11:35:41 AM UTC-5, Joao Cardoso wrote:

I think it is, the awk error is just Alt-F trying to compute the current elapsed percentage.

And you don't need to wait for completion on that page, you can go, e.g. to the Status page, then return to the Filesystem page...

As for the first (swap) partition, you have to use the command:

   mkswap /dev/sda1

   swapon /dev/sda1

ah, and you might be right, without a swap partition the creation of the filesystem might fail. If in the status page you don't see the completion percentage to increase, it is not working and you will have to repeat the procedure after creating the swap as above.

 

Is there a command line you would like me to try with this one?

Thanks!
Scott

That did the trick!
Making the swap, and turning it on, fixed that issue.

It is now definitely properly formatting it, as I can see the "mkfs.ext4" process running now.

Coming back the page, it is showing the percentage formatted as well now.
Things are looking great! :)

Scott

[Joao Cardoso]

One more thing, please. It will not destroy your partition table or filesystems.

On the now fully partitioned disk, please run

   sgdisk --set-alignment=8 --first-in-largest /dev/sda; echo $?

   sgdisk --set-alignment=8 --end-of-largest /dev/sda; echo $?

It's just to see how it reports a fully partitioned disk.

The issue it's now diagnosed and a turnaround found.

Thanks for all your collaboration, I will now fix the issue(s) for the next release.

Joao

 

[ScottK]

On Wednesday, May 2, 2012 12:09:25 PM UTC-5, Joao Cardoso wrote:

One more thing, please. It will not destroy your partition table or filesystems.

On the now fully partitioned disk, please run

   sgdisk --set-alignment=8 --first-in-largest /dev/sda; echo $?

   sgdisk --set-alignment=8 --end-of-largest /dev/sda; echo $?

It's just to see how it reports a fully partitioned disk.

The issue it's now diagnosed and a turnaround found.

Thanks for all your collaboration, I will now fix the issue(s) for the next release.

Joao

 

The results:

# sgdisk --set-alignment=8 --first-in-largest /dev/sda; echo $?
34
0

# sgdisk --set-alignment=8 --end-of-largest /dev/sda; echo $?
63
0

Thanks again for all your help, this worked out really well! :)
Scott

[Dick O]

Thanks Joao and ScottK!

Seagate Barracuda 3TB running nicely in my DNS-323 rev. C1 / Alt-F 0.1RC2.

[Andy O.]

Thank you for your help guys. Maybe its a wrong location to ask this but how can i joint two disks to JBOD.Raid utility does not list any of the drives (says note in every group) I have 2 and 3 tb disks, that i would like to use as one.

RAID Creation

fdisk: device has more than 2^32 sectors, can't use all of them

Dev.	Type	Comp 1	Comp 2	Spare
md0	JBD Raid 0 Raid 1 Raid 5	none	none	none

May be there is a way to do it via telnet?

Thanks

[Joao Cardoso]

On Thursday, May 17, 2012 5:02:37 AM UTC+1, Andy O. wrote:

Thank you for your help guys. Maybe its a wrong location to ask this

Yes, it is, please open a new topic on setting up JBOD and 3TB disks

And read first the online help available at the RAID page

[Mike Kirk]

I followed these instructions and managed to create a mirror from a pair of 3TB Seagates, Thanks! I posted my instructions based on this thread with some screenshots.

[Joao Cardoso (Alt-F)]

On Sunday 24 June 2012 14:43:11 Mike Kirk wrote:
> I followed these instructions and managed to create a mirror from a pair of
> 3TB Seagates, Thanks! I posted my

> instructions<https://www.mikekirk.org/index.php/2012/06/24/using-3tb-drives
> -in-your-d-link-dns-323/>based on this thread with some screenshots.

Nice, thanks for posting this.

But I really need to find what is wrong, so I can correct it and make cmdline
afraid people happy.
To fix it I need to diagnose it first, as I don't have (and don't need) a 3TB
disk.

With other users colaboration, I think that I now have it fixed for the next
release, but Eclipzeex follow those instructions up to a point, and now he is
stuck.
With his collaboration I hope to discover what is avoiding 'mdadm' to create
the RAID array.

Thanks!

>
> <https://www.mikekirk.org/images/computers/DNS-323/11-windows.jpg>

[Joao Cardoso]

Oops, I though I was replying to the thread "RAID-1 from 2 x 3TB WD cavier green failed" initiated by 'Eclipzeex'

Anyway, upon closer examination, there are two possible issues with your instructions:

 -the lack of an internal bitmap, which will imply a lengthily resync everytime the RAID is assembled. This can be overcome by using in the Alt-F RAID web page, 'RAID Operations', 'Add Bitmap', (or adding '--bitmap=internal' at RAID creation time)

 -by default mdadm uses RAID metadata 1.2, if other is not specified. No problem, but if you change your mind and want to go back to the D-Link firmware (and why would you do that? :-) it will not recognize the RAID. This can only be overcome at RAID creation time by adding '--metadata=0.9' as one option.

I know, I'm picky :-(

Thanks for the instructions!

[Mike Kirk]

> Anyway, upon closer examination, there are two possible issues with your instructions:

Fixed:
- Defaulted to creation of the array with internal bitmap (makes sense: losing a bit of I/O to the bitmap on an already slow device is much less important than a resync taking all day)
- Made note that the default build command won't let you return to OEM firmware, and listed your alternate 0.90-metadata option

   Mike

[Will Kallander]

Mike, your summarized instructions worked perfectly for me, on a DNS-323 v.A1 w/ Alt-F freshly installed with the new Seagate ST3000DM001 3TB HDD.

One surprise though, when I copied your commands to paste into the ssh terminal, the double hyphens were collapsed into a single character, and the bash interpreter on the NAS treated them as a control char.  Specifically, this is an example of what I got back:

# sgdisk –zap-all /dev/sda
Problem opening –zap-all for reading! Error is 2.
The specified file does not exist!

Easy to fix by search and replacing "–" with "--", I just wanted to let you know - looked like some weird thing in WordPress' editor.

So... this thread made it a snap to jump up to a 3TB on my old reliable NAS. Awesome job you guys, thanks so much!

[Joao Cardoso]

As this is one of the most read topics, and RC3 is never seeing the light of the day, I attach four files that should fix the Disk Partitioner and Disk Wizard issue with 3TB disks.

Please download the attached files to the box, then execute the commands:

cd <where you downloaded the files>

mv diskpart*.cgi diskwiz*.cgi /usr/www/cgi-bin

chmod +x /usr/www/cgi-bin/*.cgi


The files will survive until the next reboot, unless you have Alt-F packages (with Alt-F web pages) installed, in which case they will survive reboots.

Please report back success or any issue.

[TJ]

"Never seeing the light of day"?

Does this mean an RC3 or a stable will never be released? Or just not in the foreseeable future? These are just temporary fixes for now? Does this include the fschk fix mentioned elsewhere in the form?

Big fan,

T.J.

[Joao Cardoso]

On Wednesday, August 29, 2012 12:35:06 AM UTC+1, TJ wrote:

"Never seeing the light of day"?

English is not my mother language, and direct translation of idiomatic wording often means "lost in translation"  :-(

No, RC3 is only taking more time to see the day light than I would expect and desire.

Since 0.1B1, released in March 2010, two and a half year have elapsed, releases used to be done every two or three months and are now done every six months, the next one should be in September, I hope I don't miss it. This means that Alt-F is more stable now, there are less things to fix, no new things to add, and motivation is not very high.

[TJ]

"Phew" *wipes brow*

I was a little worried that the project had come to a premature end. I'm certain that I will be an early adopter of RC3 and will flash to my two 323's as soon as it releases. I meant no disrespect, and am not bothered by the time between releases.

All the best,

T.J.

[Rolf Larsson]

Joao,

I'm attempting to install 3TB barracudas in my two 323's. Should I wait for RC4 (or is there a prerelease somewhere I can try?), or just move ahead with the manual overrides?

-- Rolf

[Rolf Larsson]

Impatiently, I ended up following mikes guide above, though I created a JBD drive instead. I'll probably wait for RC4 before I upgrade the other one, though.

[Joao Cardoso]

I would appreciate if you could test the fix (attachments) that I posted a few posts above. Having feedback on this issue is crucial for me and RC3.

I would appreciate still more if you could test more disk wizard configurations than the one that you are interested in.

RC3 is delayed because there is a long standing mis-feature that I would like to fix, and I don't get the time to do it (I'm in the process of house remodeling...)

On Sep 22, 2012 8:58 PM, "Rolf Larsson" <rolf.l...@gmail.com> wrote:

Impatiently, I ended up following mikes guide above, though I created a JBD drive instead. I'll probably wait for RC4 before I upgrade the other one, though.

On Saturday, September 22, 2012 9:22:57 PM UTC+2, Rolf Larsson wrote: > > Joao, > > I'm attempting...

To view this discussion on the web visit https://groups.google.com/d/msg/alt-f/-/e_yHA5Z4eUwJ.
 
 

[Rolf Larsson]

Sadly, the unit is now being filled with data. Once done, I can play around with the other unit a bit (I have two more 3TB disks that will go into that one).

-- Rolf

[Saya]

Hi,

I followed mike guide, use my 3TB drive and got this situation:

# df -h | grep /mnt/sda2
/dev/sda2                 2.7T      2.5T         0 100% /mnt/sda2

2.7T size, 2.5T used but 0 byte available, I can not add more file using CIFS, how to fix this ? (Note: I try to mount the drive using my linux box and still can use the free space left)

thanks

[Joao Cardoso]

Mike's instructions are for a RAID setup, two disks, while you seems to only have one drive.
You can follow the instructions until the first 'mdadm' command -- you will have your disk correctely partitioned and swap active. You then have to create a filesystem on sda2, use Disk->Filesystems.

Or you could help Alt-F and test the scripts that I posted a few posts ago, those 4 files with a .cgi extension. You can then try to use Disk->Wizard that should do everything needed. If you do that, please report back your results, either success or insuccess.

-- You received this message because you are subscribed to the Google Groups "Alt-F" group. To po...

To view this discussion on the web visit https://groups.google.com/d/msg/alt-f/-/LrNYm6cGdHAJ.
 
 

[Joao Cardoso]

Ah, if you have followed Mike's instructions and have only one disk, you should now have a degraded RAID array (see the status page), and you have to destroy it first in the Disk->RAID web page (loosing all data on it) before creating a filesystem on sda2.
Read the Filesystem and RAID web pages on-line help.

> > On Sep 29, 2012 5:27 PM, "Saya" <okto...@gmail.com> wrote: > > Hi, > > I followed mike guide, ...

-- You received this message because you are subscribed to the Google Groups "Alt-F" group. To po...

> > To view this discussion on the web visit https://groups.google.com/d/msg/alt-f/-/LrNYm6cGdHAJ. ...

[Saya]

I skipped the RAID part of Mike's instruction because I don't need it. I try to mount the partition in my linux box and everything seems normal, I also copy files to /mnt/sda2 using my linux box and found no problems. If being mounted in Alt-F, I also can read its content as well, the only problem is the disk still shown as "no available space". Should I repartition and use the web UI instead ?

[Joao Cardoso]

You said that you followed, now that you skipped some instructions - which one have you followed and which one have you skipped? No mind-reader this side ;)

On Sep 30, 2012 3:54 AM, "Saya" <okto...@gmail.com> wrote:

I skipped the RAID part of Mike's instruction because I don't need it. I try to mount the partition in my linux box and everything seems normal, I also copy files to /mnt/sda2 using my linux box and found no problems. If being mounted in Alt-F, I also can read its content as well, the only problem is the disk still shown as "no available space". Should I repartition and use the web UI instead ?

On Sunday, September 30, 2012 6:14:06 AM UTC+7, Joao Cardoso wrote:

> > Ah, if you have followed Mike's instructions and have only one disk, you should now have a degr...

> > On Sep 29, 2012 5:27 PM, "Saya" <okto...@gmail.com> wrote: > > Hi, > > I followed mike guide, ...

>> >> -- You received this message because you are subscribed to the Google Groups "Alt-F" group. T...

-- You received this message because you are subscribed to the Google Groups "Alt-F" group.

To post to this group, send email to al...@googlegroups.com. To unsubscribe from this group, send em...

To view this discussion on the web visit https://groups.google.com/d/msg/alt-f/-/9vlT-ygztUQJ.
 
 

[Saya]

I'm sorry, here's complete list of what I did:

# sgdisk --zap-all /dev/sda

# sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda

# end=$(sgdisk --set-alignment=8 --end-of-largest /dev/sda)

# sgdisk --set-alignment=8 --new=2:0:${end} --typecode=2:8300 /dev/sda
# sgdisk -p /dev/sda

# mkswap /dev/sda1
# swapon /dev/sda1

and then I created new ext4 partition using web interface.

thanks

[Joao Cardoso]

On Monday, October 1, 2012 2:14:25 AM UTC+1, Saya wrote:

I'm sorry, here's complete list of what I did:

# sgdisk --zap-all /dev/sda
# sgdisk --set-alignment=8 --new=1:64:+512M --typecode=1:8200 /dev/sda
# end=$(sgdisk --set-alignment=8 --end-of-largest /dev/sda)
# sgdisk --set-alignment=8 --new=2:0:${end} --typecode=2:8300 /dev/sda
# sgdisk -p /dev/sda

# mkswap /dev/sda1
# swapon /dev/sda1

and then I created new ext4 partition using web interface.

Everything looks fine (and is a valid recipe for a "standard", non-RAID, disk layout).

I have misread your first post, whose subject should be "Can't use all space of my 3TB disk". Unfortunately Topics can't be moved.

You should also said: "I followed mike guide, [missing: starting adding files without problems to] my 3TB drive and got this situation:"

> # df -h | grep /mnt/sda2

> /dev/sda2                 2.7T      2.5T         0 100% /mnt/sda2

> 

> 2.7T size, 2.5T used but 0 byte available, I can not add more file using CIFS, how to fix this ? (Note: I try to mount the drive using

> my linux box and still can use the free space left)

> I try to mount the partition in my linux box and everything seems normal, 

> I also copy files to /mnt/sda2 using my linux box and found no problems.

> If being mounted in Alt-F, I also can read its content as well, the only problem is the disk still shown as "no available space".

This can be a bug in busybox or uclibc.

-What does 'df' shows when you use the disk on your linux box? Some available space, I guess.

 When in Alt-F:

-Can you use the command line 'cp' command to create new files, or some other command, such as 'tar'?

-Can you transfer files to the box using ftp or ssh (or using 'wget' in the 323)?

-Have you set the reserved space to 0 using 'tune2fs -m 0 /dev/sda2'?

-Can you please post the output of 'tune2fs -l /dev/sda2'? 

-Does the output of 'dumpe2fs /dev/sda2 | grep 'Free blocks''' shows some numbers, such as "Free blocks: 434433-442368"?

Thanks

[Callino]

Hi,
 
I am looking for Mike's instructions but, unfortunately, they are not available anymore under this link: https://www.mikekirk.org/2012/06/24/using-3tb-drives-in-your-d-link-dns-323/

Would be great if someone could help me and post a copy here.

Thanks
Callino

[Puru]

New URL
https://www.mikekirk.org/wordpress/using-3tb-drives-in-your-d-link-dns-323/

[Joao Cardoso]

On Wednesday, November 28, 2012 12:22:44 AM UTC, Puru wrote:

New URL
https://www.mikekirk.org/wordpress/using-3tb-drives-in-your-d-link-dns-323/

The easier way to format 3TB disks will be to use the RC3 snapshot announced in one of the top-posted topics, and TryIt; after formatting the disks you can reboot the box and start using them under RC2. DON'T flash the RC3 snapshot, instead use the TryIt button in the Firmware Upgrade page.

Unfortunately Mike's instructions have an error. It is not his fault, it's mine.

I suggested him to add instructions for using RAID metadata version 0.9, to maintain compatibility with DLink firmware, in case the user want to revert back to DLink firmware. He followed the suggestion, but it makes no sense at all, because DLink firmware does not supports 3TB disks, and, worse, RAID metadata 0.9 does not supports 3TB disks!

My apologies to Mike for inducing him in error.

Joao

[TJ]

Can I ask something quickly?

 

One of my two DNS-323s is full (again) and I currently have one 2.0TB and one 1.5TB drive running JBOD (once tried RAID-0 but saw no performance increase). With the RC3 snapshot, or in the future using a real RC3 am I able to destroy my array, format, swap out my 1.5TB with a 3TB and rebuild a JBOD, or am I going to run into issues mixing GPT and MBR disks in a raid or raid-like setup? I'm trying to hold off buying new hardware as long as possible.

 

Thanks,

 

T.J.

[Joao Cardoso]

On Thursday, November 29, 2012 1:55:23 AM UTC, TJ wrote:

Can I ask something quickly?

 

One of my two DNS-323s is full (again) and I currently have one 2.0TB and one 1.5TB drive running JBOD (once tried RAID-0 but saw no performance increase). With the RC3 snapshot, or in the future using a real RC3 am I able to destroy my array, format, swap out my 1.5TB with a 3TB and rebuild a JBOD,

A JBOD can't be rebuild, it is just one disk concatenated with the other, i.e., data will start filling the first disk and when that disk becomes full data will start filling the second disk. If you remove the first disk you loose all your data (unless you hire a linux filesystem expert); if you remove the second disk you will loose only the data that it contains.

But you said: "am I able to destroy my array",... so you are not concerned with your current data. And where you said "and rebuild a JBOD", you mean "create" a jBOD, right?

If that is the case, yes, you can destroy your current JBOB, loosing all your data, remove the 1.5TB disk, insert a new 3TB disk, and create a new JBOD. You can do that either manually using the Disk Partitioner, followed by the RAID page, followed by the Filesystem page, or just using the Disk wizard, that does it all for you.

You have to use the RC3 snapshot, as one of the disks is greater than 2.2TB

or am I going to run into issues mixing GPT and MBR disks in a raid or raid-like setup?

No problem, I have a RAID-1 and RAID-0 where one of the disks has GTP and the other has MBR (and both disks have only 80GB; GTP *has* to be used (only if you want to use all its space) on disks greater than 2.2TB, but *can* be used on smaller disks. RAID or filesystems have no idea that they are dealing with GTP or MBR)

[Scott Vincent]

Running Alt-F 0.1RC2.

This topic is of GREAT interest to me as well as I have just purchased 2 ea. WD 3TB RED drives for use in my 323 and am having the very same issues as described above.

I am using Mike's layman WP site to try and get these running as I hope they will in my RAID1 setup.



Let me know how I can be a tool for you.


Scott

[Scott Vincent]

Running Alt-F 0.1RC2..


Looks as if my WD Red 3TB HDD's cannot be configured in any kind of setup, RAID0, 1, JBOD etc when I try to use the WIzard to setup RAID1, as ext4, I am getting an error message:

Stopping all services and disks... done.

GPT partitioning disk sda.../dev/sda raid equal 5860531120 done.

GPT partitioning disk sdb.../dev/sdb raid equal 5860531120 done.

fdisk: device has more than 2^32 sectors, can't use all of them fdisk: device has more than 2^32 sectors, can't use all of them

Creating and activating swap in disk sda... done.

Creating and activating swap in disk sdb... done.

Creating raid1... failed:

Under Disk utilities i get:

right	sda	3000.6 GB	WDC WD30EFRX-68AX9N0
left	sdb	3000.6 GB	WDC WD30EFRX-68AX9N0

Under Disk Filesystems I get:

sda2	2794.0GB
sdb2	2794.0GB

 Under Disk RAID it shows me that I can select RAID1 and the drop downs show each of my 2 HDD's as selectable, sda2 - 3000Gb and sdb2 - 3000GB. When i select them both, and create RAID, i get an error message that RAID Creation Failed: with no reason.

HELP Please...


Additional item of interest, although i am able to work with the NAS vie Putty and web browser, I am not able to see it on the network... elementary issue I am sure, but I am lost at this point.

Scott

Scott

[Joao Cardoso]

On Saturday, December 1, 2012 3:57:49 PM UTC, Scott Vincent wrote:

Running Alt-F 0.1RC2..

You have to use the RC3 snapshoot, see the top-posted topics. Use TryIt, *NOT* Flash it, in System->FIrmware.

 

Looks as if my WD Red 3TB HDD's cannot be configured in any kind of setup, RAID0, 1, JBOD etc when I try to use the WIzard to setup RAID1, as ext4, I am getting an error message:

Stopping all services and disks... done.

GPT partitioning disk sda.../dev/sda raid equal 5860531120 done.

GPT partitioning disk sdb.../dev/sdb raid equal 5860531120 done.

fdisk: device has more than 2^32 sectors, can't use all of them fdisk: device has more than 2^32 sectors, can't use all of them

Creating and activating swap in disk sda... done.

Creating and activating swap in disk sdb... done.

Creating raid1... failed:

Under Disk utilities i get:

right	sda	3000.6 GB	WDC WD30EFRX-68AX9N0
left	sdb	3000.6 GB	WDC WD30EFRX-68AX9N0

Under Disk Filesystems I get:

sda2	2794.0GB
sdb2	2794.0GB

 Under Disk RAID it shows me that I can select RAID1 and the drop downs show each of my 2 HDD's as selectable, sda2 - 3000Gb and sdb2 - 3000GB. When i select them both, and create RAID, i get an error message that RAID Creation Failed: with no reason.

Again, try the RC3 snapshot. But the (RC3) Disk Wizard should do it all.

 

HELP Please...


Additional item of interest, although i am able to work with the NAS vie Putty and web browser, I am not able to see it on the network... elementary issue I am sure, but I am lost at this point.

You might have to start the samba/CIFS/smb service: Services->Network->smb->Configure/Start

But it only makes sense after having a filesystem. And users.

Joao

 

Scott

Scott

[Scott Vincent]

Thank you Sir!

I will advise on the results of my attempt.

Regards,

Scott

Sent from my iPhone

--

You received this message because you are subscribed to the Google Groups "Alt-F" group.
To post to this group, send email to al...@googlegroups.com.

To unsubscribe from this group, send email to alt-f+un...@googlegroups.com.
Visit this group at http://groups.google.com/group/alt-f?hl=en.
To view this discussion on the web visit https://groups.google.com/d/msg/alt-f/-/sEY5xOU_UOYJ.
 
 

[Saya]

Hi, sorry for very late reply.

-Have you set the reserved space to 0 using 'tune2fs -m 0 /dev/sda2'?

Thanks, after set using tune2fs, I can reclaim the free space.

[Sam Spiridonov]

Hi All,

I've just bought the same HDDs, but the commands from www.mikekirk.org are not exactly executable in RC3 (for some reasons, no idea why).

That's the latest version works well on my DNS-323 (rev C1) with 2 Seagate 3TB drives:

Putty:

sgdisk -a=8 -n=1:64:1048576 -t=1:8200 /dev/sda

sgdisk -a=8 -n=2:1048577 -t=2:8300 /dev/sda

sgdisk -p /dev/sda

sgdisk -a=8 -n=1:64:1048576 -t=1:8200 /dev/sdb

sgdisk -a=8 -n=2:1048577 -t=2:8300 /dev/sdb

sgdisk -p /dev/sdb

mkswap /dev/sda1

mkswap /dev/sdb1

swapon /dev/sda1

swapon /dev/sdb1

Then go to Alt-F and use Disk - Partitioner to change the system type for sda2 and sdb2 to 'RAID'

mdadm -C /dev/md0 --bitmap=internal -c64 --metadata=0.90 -l1 -n2 /dev/sda2 /dev/sdb2 

this command has zero output. No errors, no active md created - just silence.

mdadm -C /dev/md0 --bitmap=internal -c64 -l1 -n2 /dev/sda2 /dev/sdb2

this one works well. Probably an old metadata is not supported anymore?

cat /proc/mdstat

resync is still in process

Then as usual - formatting of md0, create samba folders for access and install transmission :)

Thank you for Alt-F, great product!

Regards,

Sam

On Tuesday, May 1, 2012 4:46:37 PM UTC+2, ScottK wrote:

Hi all,

I recently bought a new Seagate 3TB drive, model ST3000DM001:
Seagate Barracuda ST3000DM001 3TB 7200 RPM 64MB Cache SATA 6.0Gb/s 3.5" Internal Hard Drive

(Actually got it from buying the "external" USB model, and opening up the case)

When installing this into the DNS-323 running the current ALT-F version of the firmware, it sees the drive, and even is ready to format the drive for me.

But the initial starting of the format fails, with quite a weird error message.

I do NOT have the error message right now, I can/will post it tonight when I get home from work.

I did nothing to the drive, in terms of removing the exiting format/partitions on the drive before I attempted to use it in Alt-F.
(Thus, it is formatted with NTFS, with about 100 Megs of Seagate docs/utils that they ship on the drive initiall)

Has anyone gotten a 3TB drive to work on Alt-F?

If not, and there is some beta, or some way to play around to make it work, let me know.
This disk is a virgin disk, and is going in a 2nd (and unused) DNS-323, thus it is available for me to muck with, without any fear of losing any existing data.

Thanks for any/all help!
Scott

[João Cardoso]

On Tuesday, June 3, 2014 8:30:44 PM UTC+1, Sam Spiridonov wrote:

Hi All,

I've just bought the same HDDs, but the commands from www.mikekirk.org are not exactly executable in RC3 (for some reasons, no idea why).

That's the latest version works well on my DNS-323 (rev C1) with 2 Seagate 3TB drives:

Putty:

sgdisk -a=8 -n=1:64:1048576 -t=1:8200 /dev/sda

sgdisk -a=8 -n=2:1048577 -t=2:8300 /dev/sda

sgdisk -p /dev/sda

sgdisk -a=8 -n=1:64:1048576 -t=1:8200 /dev/sdb

sgdisk -a=8 -n=2:1048577 -t=2:8300 /dev/sdb

sgdisk -p /dev/sdb

mkswap /dev/sda1

mkswap /dev/sdb1

swapon /dev/sda1

swapon /dev/sdb1

Then go to Alt-F and use Disk - Partitioner to change the system type for sda2 and sdb2 to 'RAID'

mdadm -C /dev/md0 --bitmap=internal -c64 --metadata=0.90 -l1 -n2 /dev/sda2 /dev/sdb2 

this command has zero output. No errors, no active md created - just silence.

 

mdadm -C /dev/md0 --bitmap=internal -c64 -l1 -n2 /dev/sda2 /dev/sdb2

this one works well. Probably an old metadata is not supported anymore?

metadata 0.9 can't be used with bigger than 2TB disks, but an error message should be output, or at least the command exit status, obtained by typing

   echo $?

immediately after executing the 'mdadm' command.

Be warned that, if you don't specify the metadata option, mdadm will use version 1.2, the default, and that will not enable "breaking" the RAID array into "standard" filesystem, which can pose problems latter when doing some maintenance operations.

For bigger than 2TB disks, I would recommend using metadata version 1.0.