Drobom and luns

[ElliotA]

Been trying to setup our drobo pro with drobom on linux with ext3
formatted partitions.
"drobom info luns" always says they are NTFS formatted even after
running "drobom format ext3 PleaseEraseMyData" Is there some way to
tell the drobo which format the luns are? I think this is affecting
our space reclamation.

[ElliotA]

Quick follow-up. With a newly created lun setup:

root@clutch:~/drobom-dev/drobo-utils# ./drobom set lunsize 2
PleaseEraseMyData
You asked nicely, so I will set the lunsize to 2 as you requested
set lunsize to 2 TiB
Done... Drobo is likely now rebooting. In a few minutes, it will come
back with the new LUN size.

root@clutch:~/drobom-dev/drobo-utils# ./drobom info luns
---------------------------------------------------------
Drobo Name: DroboPro Devices: /dev/sdc:/dev/sdd:/dev/sde
Time: Tue Dec 15 11:15:54 2009
---------------------------------------------------------
hoho! multiple partition types! Brave are we not?
lun size (GB) used PTFmt FStype:
0 2199 0 No Partitions ['FAT32']
0 0 0 No Partitions ['NTFS']
0 0 0 GPT ['NTFS', 'EXT3']
---------------------------------------------------------

[Peter Silva]

format, well, does not format. it writes a script to /tmp, called
/tmp/fmtscript, for you to run.
It says so in it´s output. after running format run
sh -x /tmp/fmtscript to actually format the drobo.

> --
>
> You received this message because you are subscribed to the Google Groups
> "drobo-talk" group.
> To post to this group, send email to drobo...@googlegroups.com.
> To unsubscribe from this group, send email to
> drobo-talk+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/drobo-talk?hl=en.
>
>
>

[ElliotA]

Thanks, yes, I had not run the entire script this time, just for the
third lun, as previously I had run all three and they still reported
as NTFS. The third lun does report that it might be ntfs or ext3 from
what I can tell. It is actually formatted as ext3

lun size (GB) used PTFmt FStype:
0 2199 0 No Partitions ['FAT32']
0 0 0 No Partitions ['NTFS']
0 0 0 GPT ['NTFS', 'EXT3']
--------------------------------------------------------

Is that how it's supposed to show?

[RickM]

Interesting. My 1st-gen drobo is formatted w/ext3, but it shows up as
ntfs:

# drobom info luns
---------------------------------------------------------
Info about Drobo Name: Drobo disk pack Devices: /dev/sdb
---------------------------------------------------------
query luninfo result:
(0, 2199023251456, 1306709602304, 'MBR', ['NTFS'])

---------------------------------------------------------

[Elliot]

Does your Drobo recover space if you delete files? Does it report the space as recovered with drobom status?

[Darryl]

Unfortunately, I don't know. My drobo just gets filled and filled and
filled, and I have 9 LEDs lit. :-( I need to stick in another disk
to top it off to the 2TB limit.

[Peter Silva]

Darryl what file system are you using?
What lunsize are you using?

with a lunsize <= 2 TiB and a file system of (ext3, FAT32, or NTFS) it
should work fine.
Several people report that if you pick a LUNSIZE > 2 TiB, space will
not be recovered.

[Darryl]

On Dec 18, 6:46 am, Peter Silva <Peter.A.Si...@gmail.com> wrote:
> Darryl what file system are you using?
> What lunsize are you using?

ext3, 2TB LUN, latest drobo firmware (1st-gen)

Current usable space is 1.4TB (two 750GB, one 1.5TB). Nine LEDs are
on, as well as a yellow LED in the empty slot (yeah, it's close). I
have a 1TB on order, and so we'll see if the number of LEDs decrease
once I put in the new drive.

[Peter.A.Silva]

when you say it gets filled, that means you are not deleting anything?
It's real data... in other words, it's "working as designed" for you,
right?

[Darryl]

On Dec 18, 12:53 pm, "Peter.A.Silva" <informa...@gmail.com> wrote:

> when you say it gets filled, that means you are not deleting anything?
> It's real data... in other words, it's "working as designed" for you,
> right?

Oh, yes. As far as I can tell, it's working fine. I'm just mildly
curious as to why drobom reports it as NTFS when it's really formatted
as ext3. It's not a big deal.

I'm not really deleting anything, as I'm using it as a media archive:
1-6GB TV shows get periodically archived to it. I may delete the
occasional file, but the large TV files tend to obliterate any newly-
freed space. :-)

Now, I just need to figure out what to do once i hit the 2TB limit.
Do I feel lucky enough to use another ext3 LUN on the same drobo? :-/

[Peter Silva]

good:  ext3 LUNS <= 2 TIB, AFAIK, are fine.  It ought to work.

bad: I don't know that any one has specifically tested using filling up the second LUN.  I don't have enough disk to test that myself.

Everyone was always banging their heads on a single LUN > 2 TiB.

my application is like yours, storage for a mythtv box, with constant churn in television programs.

the extfs vs. ntfs.  I found one possibility.  Can you do try the version in the git repository and let me know if it's any better?

[Elliot]

We are using a drobopro with 2TB luns. We have 3 so far, all formatted ext3. We're using all 3 so far, so good, though we just went into production last week. Each lun is between 10 and 30% full.

We did not have good luck with 8 or 16TB luns.

I believe you still need to format the new luns with partitions when you add new drives, but the drobom format ext3 command will create a script for formatting all luns and you could just remove the first few and run only those you need.

[Darryl]

On Dec 19, 8:11 am, Peter Silva <informa...@gmail.com> wrote:

> the extfs vs. ntfs. I found one possibility. Can you do try the version in
> the git repository and let me know if it's any better?

No difference, although there is an extra paren on line 319 of
DroboIOctl.py.

drobom -v 8 info looks like:

DMIP sub_page query:0x01 pattern: >BBHBBBQ
DMIP sub_page query:0x01 pattern: >BBHBBBQ
DMIP response sub_page:0x01 returned: (4, 0, 16, 4294967159)
DMIP sub_page query:0x05 pattern: >BBHLH32s
DMIP response sub_page:0x05 returned: (1261272717, 2048, 'Drobo disk
pack
\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00')
DMIP sub_page query:0x08 pattern: >BBHBBHBB32s16s16s240s
DMIP response sub_page:0x08 returned: (1, 252, 21110, 14, 6, 'Aug 31
2009,18:12:03\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00',
'ArmMarvell ',
'1.3.5\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00',
'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00/\xff
\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00F
\x03\x00\x00\x01]P\xf6`\x00\x00\x00\x00\x00\x00\x00\x00\x00\x03SAMSUNG
HD154UI SAMSUNG HD154UI \x00\x00\x1d
\x00\x00\x00\x01\xff\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\xa7\xa1\xba\xe3\x8e')
DMIP sub_page query:0x01 pattern: >BBHBBBQ
DMIP response sub_page:0x01 returned: (4, 0, 16, 4294967159)
DMIP sub_page query:0x05 pattern: >BBHLH32s
DMIP response sub_page:0x05 returned: (1261272717, 2048, 'Drobo disk
pack
\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00')
DMIP sub_page query:0x08 pattern: >BBHBBHBB32s16s16s240s
DMIP response sub_page:0x08 returned: (1, 252, 21110, 14, 6, 'Aug 31
2009,18:12:03\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00',
'ArmMarvell ',
'1.3.5\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00',
'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00/\xff
\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x01\xff\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0b\x00\x00\x01\xff
\xff\xff\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0c
\x00\x00\x01\xff\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\r\x00\x00\x01\xff\xff
\xff\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0e
\x00\x00\x01\xff\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0f\x00\x00\x01\xff
\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\xe3\xb0\xa6\xa0\xac')
DMIP sub_page query:0x01 pattern: >BBHBBBQ
DMIP response sub_page:0x01 returned: (4, 0, 16, 4294967159)
(4, 16, 2199023185408)
DMIP sub_page query:0x02 pattern: >BBHQQQQ
DMIP response sub_page:0x02 returned: (131130703872, 1364474216448,
1495604920320, 374892134400)
(131130703872, 1364474216448, 1495604920320, 374892134400)
DMIP sub_page query:0x06 pattern: >BBHBB
DMIP response sub_page:0x06 returned: (0, 11)
(0, 11)
DMIP sub_page query:0x03 pattern:
>BBHBHBQQB32s16sLHBQQB32s16sLHBQQB32s16sLHBQQB32s16sL
DMIP response sub_page:0x03 returned: (4, 70, 0, 750156374016, 0, 3,
'ST3750330AS ', 'ST3750330AS ', 9984, 70, 1,
750156374016, 0, 3, 'WDC WD7500AAKS-00RBA0 ', 'WDC
WD7500AAKS-0', 10752, 70, 2, 0, 0, 130,
'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00',
'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00', 0,
70, 3, 1500301910016, 0, 3, 'SAMSUNG HD154UI ',
'SAMSUNG HD154UI ', 7424)
0 750 ST3750330AS green
1 750 WDC WD7500AAKS-00RBA0 green
2 0 yellow
3 1500 SAMSUNG HD154UI green
DMIP sub_page query:0x08 pattern: >BBHBBHBB32s16s16s240s
DMIP response sub_page:0x08 returned: (1, 252, 21110, 14, 6, 'Aug 31
2009,18:12:03\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00',
'ArmMarvell ',
'1.3.5\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00',
'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00/\xff
\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x01\xff\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0b\x00\x00\x01\xff
\xff\xff\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0c
\x00\x00\x01\xff\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\r\x00\x00\x01\xff\xff
\xff\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0e
\x00\x00\x01\xff\xff\xff
\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\x00\x11\x0f\x00\x00\x01\xff
\xff\xff\xf0\x00\x00\x00\x08\x00\x00\x00\x00\x00\xb4\xa6\xb1\x8e
\xa4')
(1, 252, 21110, 14, 6, 'Aug 31 2009,18:12:03', 'ArmMarvell', '1.3.5',
['NO_AUTO_REBOOT', 'NO_FAT32_FORMAT', 'USED_CAPACITY_FROM_HOST',
'DISKPACKSTATUS', 'ENCRYPT_NOHEADER', 'CMD_STATUS_QUERIABLE',
'VARIABLE_LUN_SIZE_1_16', 'PARTITION_LUN_GPT_MBR',
'FAT32_FORMAT_VOLNAME', 'SUPPORTS_DROBOSHARE',
'SUPPORTS_NEW_LUNINFO2', 'feature x0800', 'feature x2000 '])
DMIP sub_page query:0x02 pattern: >BBHQQQQ
DMIP response sub_page:0x02 returned: (131130703872, 1364474216448,
1495604920320, 374892134400)
DMIP sub_page query:0x05 pattern: >BBHLH32s
DMIP response sub_page:0x05 returned: (1261272717, 2048, 'Drobo disk
pack
\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00')
DMIP sub_page query:0x09 pattern: >BBHLL
DMIP response sub_page:0x09 returned: (32772, 0)
/dev/sdb /archive/big1 Drobo disk pack 91% full - (['Yellow alert'],
0)
DMIP sub_page query:0x30 pattern: >BBHBBBIBB
DMIP response sub_page:0x30 returned: (85, 95, 0, 0, 0, 0)
{'RedThreshold': 95, 'YellowThreshold': 85}
DMIP sub_page query:0x04 pattern:
>BBHBHBQQHBQQHBQQHBQQHBQQHBQQHBQQHBQQ
DMIP response sub_page:0x04 returned: (1, 17, 0, 1495604920320,
1364474216448, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
DMIP sub_page query:0x07 pattern:
>BBHBHBQBBBH3BHBQBBBH3BHBQBBBH3BHBQBBBH3BHBQBBBH3BHBQBBBH3BHBQBBBH3BHBQBBBH3B
DMIP response sub_page:0x07 returned: (1, 17, 0, 2199023251456, 1, 1,
8, 0, 0, 0, 0, 17, 1, 2199023251456, 0, 0, 8, 0, 0, 0, 0, 17, 2,
2199023251456, 0, 0, 8, 0, 0, 0, 0, 17, 3, 2199023251456, 0, 0, 8, 0,
0, 0, 0, 17, 4, 2199023251456, 0, 0, 8, 0, 0, 0, 0, 17, 5,
2199023251456, 0, 0, 8, 0, 0, 0, 0, 17, 6, 2199023251456, 0, 0, 8, 0,
0, 0, 0, 17, 7, 2199023251456, 0, 0, 8, 0, 0, 0, 0)
(0, 2199023251456, 1364474216448, 'MBR', ['NTFS'])

[Darryl]

On Dec 19, 8:11 am, Peter Silva <informa...@gmail.com> wrote:

> good:  ext3 LUNS <= 2 TIB, AFAIK, are fine.  It ought to work.

Just got my 1TB disk. I now have ~2.3TB of usable space, and, as you
said, the number of LEDs did indeed decrease (from 9 to 5). I
formatted the second LUN as a 2TB ext3 filesystem, and so we'll see
how that goes.