| View previous topic :: View next topic |
| Author |
Message |
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
 Posted: Mon Feb 06, 2012 8:55 am Post subject: 2.5TB hdd (GTP) says disk full, but 1.5TB left.. ?? |
 |
|
i am just filling my backup drive.. and am at this step:
| Code: |
/dev/sdc1 2,3T 881G 1,5T 38% /mnt/BACKUP
|
as you can see, there is 1.5TB space left. it is partitioned with parted, and has a GPT table..
now if i continue filling the disk:
| Code: |
p -v 19600000_001.jpg /mnt/BACKUP
„19600000_001.jpg“ -> „/mnt/BACKUP/19600000_001.jpg“
cp: reguläre Datei „/mnt/BACKUP/19600000_001.jpg“ kann nicht angelegt werden: Auf dem Gerät ist kein Speicherplatz mehr verfügbar
|
sorry the german.. it says there is no space left on the device, wtf??
i already copied 880GB fine.
the filesystem is ext3 with 4096 sectorsize
| Code: |
GPT fdisk (gdisk) version 0.6.13
Partition table scan:
MBR: protective
BSD: not present
APM: not present
GPT: present
Found valid GPT with protective MBR; using GPT.
Command (? for help): p
Disk /dev/sdc: 4883781168 sectors, 2.3 TiB
Logical sector size: 512 bytes
Disk identifier (GUID): 3569D057-2690-4BC1-AB1A-4777409158F2
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 4883781134
Partitions will be aligned on 2048-sector boundaries
Total free space is 3565 sectors (1.7 MiB)
Number Start (sector) End (sector) Size Code Name
1 2048 4883779583 2.3 TiB 0700 primary
|
any clues?
wtf??
| Code: |
Total free space is 3565 sectors (1.7 MiB)
|
just realized the above while posting it here.. |
|
| Back to top |
|
 |
Ant P.
Veteran
Joined: 18 Apr 2009
Posts: 1295
Location: UK
|
| Posted: Mon Feb 06, 2012 9:33 am Post subject: Re: 2.5TB hdd (GTP) says disk full, but 1.5TB left.. ?? |
|
|
| Dieter.Soltau wrote: | | the filesystem is ext3 with 4096 sectorsize |
You ran out of inodes. `df -i` will confirm this.
_________________
sig temporarily out of order |
|
| Back to top |
|
|
Kaso_da_Zmok
Apprentice
Joined: 19 Aug 2004
Posts: 176
Location: Cork, Ireland
|
| Posted: Mon Feb 06, 2012 9:43 am Post subject: |
|
|
| Quote: | Number Start (sector) End (sector) Size Code Name
1 2048 4883779583 2.3 TiB 0700 primary |
this is just cosmetic issue
but the partition type you have there is NTFS.
should look like this instead/
| Code: | Number Start (sector) End (sector) Size Code Name
1 2048 4883779583 2.3 TiB 8300 Linux filesystem
|
|
|
| Back to top |
|
|
kingfame_147
Apprentice
Joined: 11 Oct 2008
Posts: 157
|
| Posted: Mon Feb 06, 2012 6:02 pm Post subject: |
|
|
| Kaso_da_Zmok wrote: | | Quote: | Number Start (sector) End (sector) Size Code Name
1 2048 4883779583 2.3 TiB 0700 primary |
this is just cosmetic issue
but the partition type you have there is NTFS.
should look like this instead/
| Code: | Number Start (sector) End (sector) Size Code Name
1 2048 4883779583 2.3 TiB 8300 Linux filesystem
|
|
?!
My gdisk says "0700 Linux/Windows data". Ubuntu use this for their data partitions as well and so do I. 8300 isn't a known type regarding my gdisk version "GPT fdisk (gdisk) version 0.6.13" |
|
| Back to top |
|
|
Kaso_da_Zmok
Apprentice
Joined: 19 Aug 2004
Posts: 176
Location: Cork, Ireland
|
| Posted: Mon Feb 06, 2012 6:49 pm Post subject: |
|
|
what planet is the 0700 linux data 
| Code: | Command (? for help): t
Partition number (1-4): 4
Current type is 'Linux filesystem'
Hex code or GUID (L to show codes, Enter = 8300): L
0700 Microsoft basic data
8300 Linux filesystem |
Ubuntu is not different, not that it matters what the partition type is. |
|
| Back to top |
|
|
kingfame_147
Apprentice
Joined: 11 Oct 2008
Posts: 157
|
| Posted: Mon Feb 06, 2012 7:02 pm Post subject: |
|
|
The planet gdisk 0.6.13 :)
And when I installed my last Ubuntu system (2 or 3 weeks ago) it set the type to 0700, too. Like I said: gdisk 0.6.13 doesn't even list 8300 as a valid partition type.
Edit:
http://www.rodsbooks.com/gdisk/walkthrough.html
Here is the type 8300 listed. Strange. In the code-block 0700 stands for Windows Data, but then it writes 0700 as Linux/Windows-Data.
Ok, I upgraded to gdisk 0.8.2 and there is 8300 available. Can I simply change it or is there a possibilty for data loss? |
|
| Back to top |
|
|
Kaso_da_Zmok
Apprentice
Joined: 19 Aug 2004
Posts: 176
Location: Cork, Ireland
|
| Posted: Mon Feb 06, 2012 8:33 pm Post subject: |
|
|
yes you can change it anytime no data loss.
it is not important what type, and if you know what fs you have on the partition.
i was just surprised that you don't have it by the book. means if you have ext3 on it than the partition type is 83 in gpt world it is 8300.
that concept of 0700 to be Windows/Linux data is new to me and type 7 or 0700 in gpt is always NTFS. |
|
| Back to top |
|
|
kingfame_147
Apprentice
Joined: 11 Oct 2008
Posts: 157
|
| Posted: Mon Feb 06, 2012 8:41 pm Post subject: |
|
|
| Well, like I said 2 times before: Get the gdisk source (gptfdisk) 0.6.13, compile and run it. You will see no 8300 type and you will see 0700 as "Linux/Windows Data". This isn't something I made up :) |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 9:14 pm Post subject: Re: 2.5TB hdd (GTP) says disk full, but 1.5TB left.. ?? |
|
|
| Ant P. wrote: | | Dieter.Soltau wrote: | | the filesystem is ext3 with 4096 sectorsize |
You ran out of inodes. `df -i` will confirm this. |
how do i fix this ?? any chance to fix without data loss? |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 9:21 pm Post subject: |
|
|
pretty odd..
| Code: |
soltau ~ # df -i
Dateisystem Inodes IBenutzt IFrei IUse% Eingehängt auf
rootfs 11231232 483446 10747786 5% /
/dev/root 11231232 483446 10747786 5% /
rc-svcdir 990098 84 990014 1% /lib64/rc/init.d
udev 990098 1048 989050 1% /dev
shm 990098 9 990089 1% /dev/shm
/dev/sdd1 615451600 167154 615284446 1% /mnt/EXT
/dev/sda2 76114308 98960 76015348 1% /mnt/XP
/dev/sda3 73876152 203461 73672691 1% /mnt/WIN7
/dev/sdb2 0 0 0 - /mnt/SHARE
/dev/sdc1 596192 596192 0 100% /mnt/BACKUP
|
what happened to sdb2 ?? it is reiserfs..
and how to fix my sdc1 now, damnit  |
|
| Back to top |
|
|
Kaso_da_Zmok
Apprentice
Joined: 19 Aug 2004
Posts: 176
Location: Cork, Ireland
|
| Posted: Mon Feb 06, 2012 9:46 pm Post subject: |
|
|
| copy all data out and reformat wtth some normal fs. ( had 2 Franziskaner Weissbier's and half a bottle of deluxe irish creame 17%) |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 27783
Location: 56N 3W
|
| Posted: Mon Feb 06, 2012 9:52 pm Post subject: |
|
|
Dieter.Soltau,
The inode count cannot be changed without destroying the filesystem.
mke2fs guesses how many inodes to make based on the filesystem size.
Read its man page - you can control the number of inodes. Either by setting a number or a ratio of blocks of storage per inode.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
Ant P.
Veteran
Joined: 18 Apr 2009
Posts: 1295
Location: UK
|
| Posted: Mon Feb 06, 2012 9:53 pm Post subject: |
|
|
| Dieter.Soltau wrote: | what happened to sdb2 ?? it is reiserfs..
and how to fix my sdc1 now, damnit :-( |
reiserfs can't run out of inodes because it doesn't use them.
_________________
sig temporarily out of order |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 9:55 pm Post subject: |
|
|
| Kaso_da_Zmok wrote: | | copy all data out and reformat wtth some normal fs. ( had 2 Franziskaner Weissbier's and half a bottle of deluxe irish creame 17%) |
i like.. ^^
well, it is a GPT drive, 4096 sectors, i am pretty unsure how to format / partition this beats at all, hmm.. |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 9:57 pm Post subject: |
|
|
| hmmmmmmmmmmm... okay.. should i use REISERFS (3) on my 2.5TB GPT drive? again: it has 4096 physical clustersize.. how the fsck should i handle that beast? |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 9:58 pm Post subject: |
|
|
| NeddySeagoon wrote: | Dieter.Soltau,
The inode count cannot be changed without destroying the filesystem.
mke2fs guesses how many inodes to make based on the filesystem size.
Read its man page - you can control the number of inodes. Either by setting a number or a ratio of blocks of storage per inode. |
okay, so whats best guess for me? it is a BACKUP drive, dunno in advance how many inodes i need.. what do you suggest for parted and filesystem? |
|
| Back to top |
|
|
Kaso_da_Zmok
Apprentice
Joined: 19 Aug 2004
Posts: 176
Location: Cork, Ireland
|
| Posted: Mon Feb 06, 2012 10:06 pm Post subject: |
|
|
Filesystems (such as JFS, or XFS) escape this limitation with extents and/or dynamic inode allocation, which can 'grow' the filesystem and/or increase the number of inodes.
my favorite is XFS |
|
| Back to top |
|
|
Ant P.
Veteran
Joined: 18 Apr 2009
Posts: 1295
Location: UK
|
| Posted: Mon Feb 06, 2012 10:07 pm Post subject: |
|
|
596k inodes seems oddly low for a 2.5TB filesystem. Using ext4 with extents would lower the usage, but it looks like mke2fs picked a bad default there. That's 1 inode per ~4MB which is an enormous waste for small files.
It would be useful for us to see the output of:
| Code: | | tune2fs -l /dev/sdc1 |
_________________
sig temporarily out of order |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 10:11 pm Post subject: |
|
|
| Ant P. wrote: | 596k inodes seems oddly low for a 2.5TB filesystem. Using ext4 with extents would lower the usage, but it looks like mke2fs picked a bad default there. That's 1 inode per ~4MB which is an enormous waste for small files.
It would be useful for us to see the output of:
| Code: | | tune2fs -l /dev/sdc1 |
|
| Code: |
soltau duke # tune2fs -l /dev/sdc1
tune2fs 1.41.14 (22-Dec-2010)
Filesystem volume name: <none>
Last mounted on: <not available>
Filesystem UUID: a226e022-6e63-41e0-81e3-d9201bf0474a
Filesystem magic number: 0xEF53
Filesystem revision #: 1 (dynamic)
Filesystem features: has_journal ext_attr resize_inode dir_index filetype needs_recovery sparse_super large_file
Filesystem flags: signed_directory_hash
Default mount options: (none)
Filesystem state: clean with errors
Errors behavior: Continue
Filesystem OS type: Linux
Inode count: 596192
Block count: 610472192
Reserved block count: 0
Free blocks: 379522027
Free inodes: 0
First block: 0
Block size: 4096
Fragment size: 4096
Reserved GDT blocks: 878
Blocks per group: 32768
Fragments per group: 32768
Inodes per group: 32
Inode blocks per group: 2
RAID stride: 1
RAID stripe width: 1
Filesystem created: Sat Feb 4 21:33:28 2012
Last mount time: Mon Feb 6 09:40:59 2012
Last write time: Mon Feb 6 09:40:59 2012
Mount count: 9
Maximum mount count: 33
Last checked: Sat Feb 4 21:33:28 2012
Check interval: 15552000 (6 months)
Next check after: Thu Aug 2 22:33:28 2012
Reserved blocks uid: 0 (user root)
Reserved blocks gid: 0 (group root)
First inode: 11
Inode size: 256
Required extra isize: 28
Desired extra isize: 28
Journal inode: 8
Default directory hash: half_md4
Directory Hash Seed: 9980ad06-d23c-4e17-8528-30166b04da8f
Journal backup: inode blocks
soltau duke #
|
btw.. just emptying the disk.. ready for suggestions. |
|
| Back to top |
|
|
Ant P.
Veteran
Joined: 18 Apr 2009
Posts: 1295
Location: UK
|
| Posted: Mon Feb 06, 2012 10:19 pm Post subject: |
|
|
1 inode per 1024 blocks... yep, the numbers add up.
I'd suggest ext4 using this:
| Code: | | mkfs.ext4 -T huge /dev/sdc1 |
That will use 64kB per inode, which will hold about 20 million files. That should be plenty (the average system has about 1 million total)
_________________
sig temporarily out of order |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 10:22 pm Post subject: |
|
|
| Ant P. wrote: | 1 inode per 1024 blocks... yep, the numbers add up.
I'd suggest ext4 using this:
| Code: | | mkfs.ext4 -T huge /dev/sdc1 |
That will use 64kB per inode, which will hold about 20 million files. That should be plenty (the average system has about 1 million total) |
what about the physical 4096k clustersize of the drive??? do i not have to specify this? i thought the drive might get slow if we use the default 512bytes??
http://www.ibm.com/developerworks/linux/library/l-4kb-sector-disks/ |
|
| Back to top |
|
|
Ant P.
Veteran
Joined: 18 Apr 2009
Posts: 1295
Location: UK
|
| Posted: Mon Feb 06, 2012 10:26 pm Post subject: |
|
|
ext4 defaults to blocksize=4096
_________________
sig temporarily out of order |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 10:30 pm Post subject: |
|
|
| Ant P. wrote: | | ext4 defaults to blocksize=4096 |
sounds good.. ty!
now what about the partitioning? i partitioned using parted, but the outcome told me 512 logical, 4096 physical.. is this okay?
i am still unsure if i partitioned the beast correctly.. i am fairly new to GPT and those new disks.. hmm..
now is the time to do it correctly so any hints welcome... |
|
| Back to top |
|
|
Dieter.Soltau
n00b
Joined: 27 Nov 2011
Posts: 54
|
| Posted: Mon Feb 06, 2012 10:49 pm Post subject: |
|
|
| Kaso_da_Zmok wrote: | Filesystems (such as JFS, or XFS) escape this limitation with extents and/or dynamic inode allocation, which can 'grow' the filesystem and/or increase the number of inodes.
my favorite is XFS |
XFS sounds good too..
any hints how to format using 4096 instead of 512?? |
|
| Back to top |
|
|
Kaso_da_Zmok
Apprentice
Joined: 19 Aug 2004
Posts: 176
Location: Cork, Ireland
|
| Posted: Mon Feb 06, 2012 10:51 pm Post subject: |
|
|
| why parted? ppl use gdisk |
|
| Back to top |
|
|
|
Other Things Gentoo
