A bug with multiple devices?

A bug with multiple devices?

[Shen Feng]

Hi,

I create a raid-1 btrfs with the following command.

[root@localhost ~]# fdisk -l /dev/sdb

Disk /dev/sdb: 320.0 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xb44829ea

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1               1        4863    39062016   83  Linux
/dev/sdb2            4864        9726    39062047+  83  Linux
/dev/sdb3            9727       38905   234375905+   f  W95 Ext'd (LBA)
/dev/sdb5            9727       14590    39063374   83  Linux
/dev/sdb6           14590       19453    39068690+   7  HPFS/NTFS
/dev/sdb7           19454       29178    78116031   83  Linux
/dev/sdb8           29179       31610    19535008+  83  Linux
/dev/sdb9           31611       34041    19526976   83  Linux
/dev/sdb10          34042       36473    19535008+  83  Linux
/dev/sdb11          36474       38904    19526976   83  Linux

[root@localhost ~]# mkfs.btrfs -m raid1 /dev/sdb8 /dev/sdb9 /dev/sdb10 /dev/sdb11
adding device /dev/sdb9 id 2
adding device /dev/sdb10 id 3
adding device /dev/sdb11 id 4
fs created label (null) on /dev/sdb8
	nodesize 4096 leafsize 4096 sectorsize 4096 size 74.50GB
Btrfs v0.16-29-g0d53b21

[root@localhost ~]# mount -t btrfs /dev/sdb8 /mnt/btrfs[root@localhost ~]# df
Filesystem           1K-blocks      Used Available Use% Mounted on
...
/dev/sdb8             78123968        28  78123940   1% /mnt/btrfs

I think the partition size is no correct since this is a raid-1 partition.
The size should be half.
Is this a bug?

-Shen

Re: A bug with multiple devices?

[Chris Samuel]

[-- Attachment #1: Type: text/plain, Size: 903 bytes --]

On Tue, 30 Dec 2008 12:44:38 pm Shen Feng wrote:

> I think the partition size is no correct since this is a raid-1 partition.
> The size should be half.

I've just run into this myself and so can confirm this issue on a btrfs 
filesystem with 2 devices, where it is created by doing:

mkfs.btrfs -m raid1 /dev/sda10 /dev/sda11

modprobe btrfs

btrfsctl -a

mount /dev/sda10 /home

The df command shows it has twice the space I was expecting (61GB rather than 
30GB).  The btrfs-show command shows that it has two devices but doesn't 
appear to contain any way to show the RAID level implemented.    I get the 
same outcome if I mount /dev/sda11 instead of sda10.

Any ideas ?

Chris
-- 
 Chris Samuel  :  http://www.csamuel.org/  :  Melbourne, VIC

This email may come with a PGP signature as a file. Do not panic.
For more info see: http://en.wikipedia.org/wiki/OpenPGP


[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 481 bytes --]

Re: A bug with multiple devices?

[Chris Samuel]

[-- Attachment #1: Type: text/plain, Size: 745 bytes --]

On Tue, 30 Dec 2008 9:45:22 pm Chris Samuel wrote:

> mkfs.btrfs -m raid1 /dev/sda10 /dev/sda11

Even doing what appears to be the correct method of:

mkfs.btrfs -d raid1 -m raid1 /dev/sda10 /dev/sda11

doesn't give any love, mkfs.btrfs still tells me the filesystem will be 60GB 
(and the kernel agrees when it is mounted).

But mkfs.btrfs does correctly set data_profile to include 
BTRFS_BLOCK_GROUP_RAID1 and does indeed call create_one_raid_group() with 
BTRFS_BLOCK_GROUP_DATA in create_raid_groups() in mkfs.c.

cheers,
Chris
-- 
 Chris Samuel  :  http://www.csamuel.org/  :  Melbourne, VIC

This email may come with a PGP signature as a file. Do not panic.
For more info see: http://en.wikipedia.org/wiki/OpenPGP


[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 481 bytes --]

Re: A bug with multiple devices?

[Yan Zheng]

2008/12/30 Chris Samuel <chris@csamuel.org>:
> On Tue, 30 Dec 2008 12:44:38 pm Shen Feng wrote:
>
>> I think the partition size is no correct since this is a raid-1 partition.
>> The size should be half.
>
> I've just run into this myself and so can confirm this issue on a btrfs
> filesystem with 2 devices, where it is created by doing:
>
> mkfs.btrfs -m raid1 /dev/sda10 /dev/sda11
>
> modprobe btrfs
>
> btrfsctl -a
>
> mount /dev/sda10 /home
>
> The df command shows it has twice the space I was expecting (61GB rather than
> 30GB).  The btrfs-show command shows that it has two devices but doesn't
> appear to contain any way to show the RAID level implemented.    I get the
> same outcome if I mount /dev/sda11 instead of sda10.
>
> Any ideas ?
>

This is a known bug. In btrfs, chunks are created dynamically and may have
different RAID configurations.  So it's difficult to compute the space
in advance.

Regards
Yan Zheng

Re: A bug with multiple devices?

[Chris Samuel]

[-- Attachment #1: Type: text/plain, Size: 509 bytes --]

On Tue, 30 Dec 2008 10:35:17 pm Yan Zheng wrote:

> This is a known bug. In btrfs, chunks are created dynamically and may have
> different RAID configurations.  So it's difficult to compute the space
> in advance.

Thanks Yan!

So mkfs.btrfs -d raid1 just sets the default value ?

All the best,
Chris
-- 
 Chris Samuel  :  http://www.csamuel.org/  :  Melbourne, VIC

This email may come with a PGP signature as a file. Do not panic.
For more info see: http://en.wikipedia.org/wiki/OpenPGP


[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 481 bytes --]

Re: A bug with multiple devices?

[Chris Samuel]

[-- Attachment #1: Type: text/plain, Size: 668 bytes --]

On Tue, 30 Dec 2008 10:35:17 pm Yan Zheng wrote:

> So it's difficult to compute the space in advance.

One thought would be to choose the most conservative RAID configuration for the 
number of devices in the system.   But then you have the opposite issue of 
potentially showing too little space in some situations.

But my guess is that for most (>50%) people once they set a default RAID level 
then most will have everything in that state.

cheers,
Chris
-- 
 Chris Samuel  :  http://www.csamuel.org/  :  Melbourne, VIC

This email may come with a PGP signature as a file. Do not panic.
For more info see: http://en.wikipedia.org/wiki/OpenPGP


[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 481 bytes --]

Re: A bug with multiple devices?

[Chris Mason]

On Tue, 2008-12-30 at 22:14 +1100, Chris Samuel wrote:
> On Tue, 30 Dec 2008 9:45:22 pm Chris Samuel wrote:
> 
> > mkfs.btrfs -m raid1 /dev/sda10 /dev/sda11
> 
> Even doing what appears to be the correct method of:
> 
> mkfs.btrfs -d raid1 -m raid1 /dev/sda10 /dev/sda11
> 
> doesn't give any love, mkfs.btrfs still tells me the filesystem will be 60GB 
> (and the kernel agrees when it is mounted).
> 
> But mkfs.btrfs does correctly set data_profile to include 
> BTRFS_BLOCK_GROUP_RAID1 and does indeed call create_one_raid_group() with 
> BTRFS_BLOCK_GROUP_DATA in create_raid_groups() in mkfs.c.

This is one of those things I've left for later.

Basically df tells you the total storage in the FS.  This storage might
be raid0, raid1 or single spindle duplication, so the usable space may
be different from the total storage.

In the future we'll have the data sent to df give better estimates of
usable space.

Changing btrfs-show to show the chunks too is a good idea.

-chris

Re: A bug with multiple devices?

[Shen Feng]

on 12/30/2008 09:44 AM, Shen Feng wrote:
> Hi,
> 
> I create a raid-1 btrfs with the following command.
> 
> [root@localhost ~]# fdisk -l /dev/sdb
> 
> Disk /dev/sdb: 320.0 GB, 320072933376 bytes
> 255 heads, 63 sectors/track, 38913 cylinders
> Units = cylinders of 16065 * 512 = 8225280 bytes
> Disk identifier: 0xb44829ea
> 
>    Device Boot      Start         End      Blocks   Id  System
> /dev/sdb1               1        4863    39062016   83  Linux
> /dev/sdb2            4864        9726    39062047+  83  Linux
> /dev/sdb3            9727       38905   234375905+   f  W95 Ext'd (LBA)
> /dev/sdb5            9727       14590    39063374   83  Linux
> /dev/sdb6           14590       19453    39068690+   7  HPFS/NTFS
> /dev/sdb7           19454       29178    78116031   83  Linux
> /dev/sdb8           29179       31610    19535008+  83  Linux
> /dev/sdb9           31611       34041    19526976   83  Linux
> /dev/sdb10          34042       36473    19535008+  83  Linux
> /dev/sdb11          36474       38904    19526976   83  Linux
> 
> [root@localhost ~]# mkfs.btrfs -m raid1 /dev/sdb8 /dev/sdb9 /dev/sdb10 /dev/sdb11
> adding device /dev/sdb9 id 2
> adding device /dev/sdb10 id 3
> adding device /dev/sdb11 id 4
> fs created label (null) on /dev/sdb8
> 	nodesize 4096 leafsize 4096 sectorsize 4096 size 74.50GB
> Btrfs v0.16-29-g0d53b21
> 
> [root@localhost ~]# mount -t btrfs /dev/sdb8 /mnt/btrfs[root@localhost ~]# df
> Filesystem           1K-blocks      Used Available Use% Mounted on
> ...
> /dev/sdb8             78123968        28  78123940   1% /mnt/btrfs
> 

More test:

[root@localhost btrfs]# df -h /dev/sdb8
Filesystem            Size  Used Avail Use% Mounted on
/dev/sdb8              75G   28K   75G   1% /mnt/btrfs
[root@localhost btrfs]# dd if=/dev/zero of=/mnt/btrfs/file bs=1024 count=1024
1024+0 records in
1024+0 records out
1048576 bytes (1.0 MB) copied, 0.0481317 s, 21.8 MB/s
[root@localhost btrfs]# sync
[root@localhost btrfs]# df -h /dev/sdb8
Filesystem            Size  Used Avail Use% Mounted on
/dev/sdb8              75G  1.1M   75G   1% /mnt/btrfs

With raid-1, the used space may be double (about 2.2M), I think.



> I think the partition size is no correct since this is a raid-1 partition.
> The size should be half.
> Is this a bug?
> 
> -Shen
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 

-- 
Best Regards,
--------------------------------------------------
Shen Feng
Development Dept.I
Nanjing Fujitsu Nanda Software Tech. Co., Ltd.(FNST)
8/F., Civil Defense Building, No.189 Guangzhou Road,
Nanjing, 210029, China
PHONE: +86-25-86630566-950
COINS: 79955-950
FAX: +86-25-83317685
MAIL: shen@cn.fujitsu.com
-------------------------------------------------
This communication is for use by the intended recipient(s) only and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not an intended recipient of this communication, you are hereby notified that any dissemination, distribution or copying hereof is strictly prohibited.  If you have received this communication in error, please notify me by reply e-mail, permanently delete this communication from your system, and destroy any hard copies you may have printed.

Re: A bug with multiple devices?

[Liu Hui]

mkfs.btrfs -m raid1 just means 'meta data' works as raid1

You should try mkfs.btrfs -d raid1

This maybe help:
mkfs.btrfs [-A <offset>] [-b <size>] [-l <size>] [-n <size>] [-l
<label>] [-s <size>] [-m raid0|raid1|raid10|single] [-d
raid0|raid1|raid10|single] <device>

2008/12/31 Shen Feng <shen@cn.fujitsu.com>:
>
>
> on 12/30/2008 09:44 AM, Shen Feng wrote:
>> Hi,
>>
>> I create a raid-1 btrfs with the following command.
>>
>> [root@localhost ~]# fdisk -l /dev/sdb
>>
>> Disk /dev/sdb: 320.0 GB, 320072933376 bytes
>> 255 heads, 63 sectors/track, 38913 cylinders
>> Units = cylinders of 16065 * 512 = 8225280 bytes
>> Disk identifier: 0xb44829ea
>>
>>    Device Boot      Start         End      Blocks   Id  System
>> /dev/sdb1               1        4863    39062016   83  Linux
>> /dev/sdb2            4864        9726    39062047+  83  Linux
>> /dev/sdb3            9727       38905   234375905+   f  W95 Ext'd (LBA)
>> /dev/sdb5            9727       14590    39063374   83  Linux
>> /dev/sdb6           14590       19453    39068690+   7  HPFS/NTFS
>> /dev/sdb7           19454       29178    78116031   83  Linux
>> /dev/sdb8           29179       31610    19535008+  83  Linux
>> /dev/sdb9           31611       34041    19526976   83  Linux
>> /dev/sdb10          34042       36473    19535008+  83  Linux
>> /dev/sdb11          36474       38904    19526976   83  Linux
>>
>> [root@localhost ~]# mkfs.btrfs -m raid1 /dev/sdb8 /dev/sdb9 /dev/sdb10 /dev/sdb11
>> adding device /dev/sdb9 id 2
>> adding device /dev/sdb10 id 3
>> adding device /dev/sdb11 id 4
>> fs created label (null) on /dev/sdb8
>>       nodesize 4096 leafsize 4096 sectorsize 4096 size 74.50GB
>> Btrfs v0.16-29-g0d53b21
>>
>> [root@localhost ~]# mount -t btrfs /dev/sdb8 /mnt/btrfs[root@localhost ~]# df
>> Filesystem           1K-blocks      Used Available Use% Mounted on
>> ...
>> /dev/sdb8             78123968        28  78123940   1% /mnt/btrfs
>>
>
> More test:
>
> [root@localhost btrfs]# df -h /dev/sdb8
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/sdb8              75G   28K   75G   1% /mnt/btrfs
> [root@localhost btrfs]# dd if=/dev/zero of=/mnt/btrfs/file bs=1024 count=1024
> 1024+0 records in
> 1024+0 records out
> 1048576 bytes (1.0 MB) copied, 0.0481317 s, 21.8 MB/s
> [root@localhost btrfs]# sync
> [root@localhost btrfs]# df -h /dev/sdb8
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/sdb8              75G  1.1M   75G   1% /mnt/btrfs
>
> With raid-1, the used space may be double (about 2.2M), I think.
>
>
>
>> I think the partition size is no correct since this is a raid-1 partition.
>> The size should be half.
>> Is this a bug?
>>
>> -Shen
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
>>
>
> --
> Best Regards,
> --------------------------------------------------
> Shen Feng
> Development Dept.I
> Nanjing Fujitsu Nanda Software Tech. Co., Ltd.(FNST)
> 8/F., Civil Defense Building, No.189 Guangzhou Road,
> Nanjing, 210029, China
> PHONE: +86-25-86630566-950
> COINS: 79955-950
> FAX: +86-25-83317685
> MAIL: shen@cn.fujitsu.com
> -------------------------------------------------
> This communication is for use by the intended recipient(s) only and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not an intended recipient of this communication, you are hereby notified that any dissemination, distribution or copying hereof is strictly prohibited.  If you have received this communication in error, please notify me by reply e-mail, permanently delete this communication from your system, and destroy any hard copies you may have printed.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>



-- 
Thanks & Best Regards
Liu Hui
--

Re: A bug with multiple devices?

[Shen Feng]

on 12/31/2008 09:17 AM, Liu Hui wrote:
> mkfs.btrfs -m raid1 just means 'meta data' works as raid1
> 
> You should try mkfs.btrfs -d raid1
> 
> This maybe help:
> mkfs.btrfs [-A <offset>] [-b <size>] [-l <size>] [-n <size>] [-l
> <label>] [-s <size>] [-m raid0|raid1|raid10|single] [-d
> raid0|raid1|raid10|single] <device>

Thanks.

The problem still exists with -d option.

[root@localhost ~]# mkfs.btrfs -m raid1 -d raid1 /dev/sdb8 /dev/sdb9 /dev/sdb10 /dev/sdb11
adding device /dev/sdb9 id 2
adding device /dev/sdb10 id 3
adding device /dev/sdb11 id 4
fs created label (null) on /dev/sdb8
	nodesize 4096 leafsize 4096 sectorsize 4096 size 74.50GB
Btrfs v0.16-29-g0d53b21
[root@localhost ~]# btrfsctl -a
Scanning for Btrfs filesystems
[root@localhost ~]#  mount -t btrfs /dev/sdb8 /mnt/btrfs
[root@localhost ~]# df -h /dev/sdb8
Filesystem            Size  Used Avail Use% Mounted on
/dev/sdb8              75G   28K   75G   1% /mnt/btrfs
[root@localhost ~]# dd if=/dev/zero of=/mnt/btrfs/file bs=1024 count=1024
1024+0 records in
1024+0 records out
1048576 bytes (1.0 MB) copied, 0.0199063 s, 52.7 MB/s
[root@localhost ~]# sync
[root@localhost ~]# df -h /dev/sdb8
Filesystem            Size  Used Avail Use% Mounted on
/dev/sdb8              75G  1.1M   75G   1% /mnt/btrfs



> 
> 2008/12/31 Shen Feng <shen@cn.fujitsu.com>:
>>
>> on 12/30/2008 09:44 AM, Shen Feng wrote:
>>> Hi,
>>>
>>> I create a raid-1 btrfs with the following command.
>>>
>>> [root@localhost ~]# fdisk -l /dev/sdb
>>>
>>> Disk /dev/sdb: 320.0 GB, 320072933376 bytes
>>> 255 heads, 63 sectors/track, 38913 cylinders
>>> Units = cylinders of 16065 * 512 = 8225280 bytes
>>> Disk identifier: 0xb44829ea
>>>
>>>    Device Boot      Start         End      Blocks   Id  System
>>> /dev/sdb1               1        4863    39062016   83  Linux
>>> /dev/sdb2            4864        9726    39062047+  83  Linux
>>> /dev/sdb3            9727       38905   234375905+   f  W95 Ext'd (LBA)
>>> /dev/sdb5            9727       14590    39063374   83  Linux
>>> /dev/sdb6           14590       19453    39068690+   7  HPFS/NTFS
>>> /dev/sdb7           19454       29178    78116031   83  Linux
>>> /dev/sdb8           29179       31610    19535008+  83  Linux
>>> /dev/sdb9           31611       34041    19526976   83  Linux
>>> /dev/sdb10          34042       36473    19535008+  83  Linux
>>> /dev/sdb11          36474       38904    19526976   83  Linux
>>>
>>> [root@localhost ~]# mkfs.btrfs -m raid1 /dev/sdb8 /dev/sdb9 /dev/sdb10 /dev/sdb11
>>> adding device /dev/sdb9 id 2
>>> adding device /dev/sdb10 id 3
>>> adding device /dev/sdb11 id 4
>>> fs created label (null) on /dev/sdb8
>>>       nodesize 4096 leafsize 4096 sectorsize 4096 size 74.50GB
>>> Btrfs v0.16-29-g0d53b21
>>>
>>> [root@localhost ~]# mount -t btrfs /dev/sdb8 /mnt/btrfs[root@localhost ~]# df
>>> Filesystem           1K-blocks      Used Available Use% Mounted on
>>> ...
>>> /dev/sdb8             78123968        28  78123940   1% /mnt/btrfs
>>>
>> More test:
>>
>> [root@localhost btrfs]# df -h /dev/sdb8
>> Filesystem            Size  Used Avail Use% Mounted on
>> /dev/sdb8              75G   28K   75G   1% /mnt/btrfs
>> [root@localhost btrfs]# dd if=/dev/zero of=/mnt/btrfs/file bs=1024 count=1024
>> 1024+0 records in
>> 1024+0 records out
>> 1048576 bytes (1.0 MB) copied, 0.0481317 s, 21.8 MB/s
>> [root@localhost btrfs]# sync
>> [root@localhost btrfs]# df -h /dev/sdb8
>> Filesystem            Size  Used Avail Use% Mounted on
>> /dev/sdb8              75G  1.1M   75G   1% /mnt/btrfs
>>
>> With raid-1, the used space may be double (about 2.2M), I think.
>>
>>
>>
>>> I think the partition size is no correct since this is a raid-1 partition.
>>> The size should be half.
>>> Is this a bug?
>>>
>>> -Shen
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>
>>>
>> --
>> Best Regards,
>> --------------------------------------------------
>> Shen Feng
>> Development Dept.I
>> Nanjing Fujitsu Nanda Software Tech. Co., Ltd.(FNST)
>> 8/F., Civil Defense Building, No.189 Guangzhou Road,
>> Nanjing, 210029, China
>> PHONE: +86-25-86630566-950
>> COINS: 79955-950
>> FAX: +86-25-83317685
>> MAIL: shen@cn.fujitsu.com
>> -------------------------------------------------
>> This communication is for use by the intended recipient(s) only and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not an intended recipient of this communication, you are hereby notified that any dissemination, distribution or copying hereof is strictly prohibited.  If you have received this communication in error, please notify me by reply e-mail, permanently delete this communication from your system, and destroy any hard copies you may have printed.
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
> 
> 
> 

-- 
Best Regards,
--------------------------------------------------
Shen Feng
Development Dept.I
Nanjing Fujitsu Nanda Software Tech. Co., Ltd.(FNST)
8/F., Civil Defense Building, No.189 Guangzhou Road,
Nanjing, 210029, China
PHONE: +86-25-86630566-950
COINS: 79955-950
FAX: +86-25-83317685
MAIL: shen@cn.fujitsu.com
-------------------------------------------------
This communication is for use by the intended recipient(s) only and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not an intended recipient of this communication, you are hereby notified that any dissemination, distribution or copying hereof is strictly prohibited.  If you have received this communication in error, please notify me by reply e-mail, permanently delete this communication from your system, and destroy any hard copies you may have printed.

Re: A bug with multiple devices?

[Shen Feng]

on 12/31/2008 09:17 AM, Liu Hui wrote:
> mkfs.btrfs -m raid1 just means 'meta data' works as raid1
> 
> You should try mkfs.btrfs -d raid1
> 
> This maybe help:
> mkfs.btrfs [-A <offset>] [-b <size>] [-l <size>] [-n <size>] [-l
> <label>] [-s <size>] [-m raid0|raid1|raid10|single] [-d
> raid0|raid1|raid10|single] <device>

Thanks.
I update the usage message of mkfs.btrfs with the following patch.

From: Shen Feng <shen@cn.fujitsu.com>

Improve the the mkfs.btrfs usage message according to
http://btrfs.wiki.kernel.org/index.php/Mkfs.btrfs

Signed-off-by: Shen Feng <shen@cn.fujitsu.com>
---
 mkfs.c |    6 +++++-
 1 files changed, 5 insertions(+), 1 deletions(-)

diff --git a/mkfs.c b/mkfs.c
index 89de8ac..be93aaa 100644
--- a/mkfs.c
+++ b/mkfs.c
@@ -263,9 +263,13 @@ static void print_usage(void)
 {
 	fprintf(stderr, "usage: mkfs.btrfs [options] dev [ dev ... ]\n");
 	fprintf(stderr, "options:\n");
+	fprintf(stderr, "\t -A --alloc-start the offset to start the FS\n");
 	fprintf(stderr, "\t -b --byte-count total number of bytes in the FS\n");
+	fprintf(stderr, "\t -d --data data profile, raid0, raid1, raid10 or single\n");
 	fprintf(stderr, "\t -l --leafsize size of btree leaves\n");
-	fprintf(stderr, "\t -n --nodesize size of btree leaves\n");
+	fprintf(stderr, "\t -L --label set a label\n");
+	fprintf(stderr, "\t -m --metadata metadata profile, values like data profile\n");
+	fprintf(stderr, "\t -n --nodesize size of btree nodes\n");
 	fprintf(stderr, "\t -s --sectorsize min block allocation\n");
 	fprintf(stderr, "%s\n", BTRFS_BUILD_VERSION);
 	exit(1);
-- 1.6.0.6 


> 
> 2008/12/31 Shen Feng <shen@cn.fujitsu.com>:
>>
>> on 12/30/2008 09:44 AM, Shen Feng wrote:
>>> Hi,
>>>
>>> I create a raid-1 btrfs with the following command.
>>>
>>> [root@localhost ~]# fdisk -l /dev/sdb
>>>
>>> Disk /dev/sdb: 320.0 GB, 320072933376 bytes
>>> 255 heads, 63 sectors/track, 38913 cylinders
>>> Units = cylinders of 16065 * 512 = 8225280 bytes
>>> Disk identifier: 0xb44829ea
>>>
>>>    Device Boot      Start         End      Blocks   Id  System
>>> /dev/sdb1               1        4863    39062016   83  Linux
>>> /dev/sdb2            4864        9726    39062047+  83  Linux
>>> /dev/sdb3            9727       38905   234375905+   f  W95 Ext'd (LBA)
>>> /dev/sdb5            9727       14590    39063374   83  Linux
>>> /dev/sdb6           14590       19453    39068690+   7  HPFS/NTFS
>>> /dev/sdb7           19454       29178    78116031   83  Linux
>>> /dev/sdb8           29179       31610    19535008+  83  Linux
>>> /dev/sdb9           31611       34041    19526976   83  Linux
>>> /dev/sdb10          34042       36473    19535008+  83  Linux
>>> /dev/sdb11          36474       38904    19526976   83  Linux
>>>
>>> [root@localhost ~]# mkfs.btrfs -m raid1 /dev/sdb8 /dev/sdb9 /dev/sdb10 /dev/sdb11
>>> adding device /dev/sdb9 id 2
>>> adding device /dev/sdb10 id 3
>>> adding device /dev/sdb11 id 4
>>> fs created label (null) on /dev/sdb8
>>>       nodesize 4096 leafsize 4096 sectorsize 4096 size 74.50GB
>>> Btrfs v0.16-29-g0d53b21
>>>
>>> [root@localhost ~]# mount -t btrfs /dev/sdb8 /mnt/btrfs[root@localhost ~]# df
>>> Filesystem           1K-blocks      Used Available Use% Mounted on
>>> ...
>>> /dev/sdb8             78123968        28  78123940   1% /mnt/btrfs
>>>
>> More test:
>>
>> [root@localhost btrfs]# df -h /dev/sdb8
>> Filesystem            Size  Used Avail Use% Mounted on
>> /dev/sdb8              75G   28K   75G   1% /mnt/btrfs
>> [root@localhost btrfs]# dd if=/dev/zero of=/mnt/btrfs/file bs=1024 count=1024
>> 1024+0 records in
>> 1024+0 records out
>> 1048576 bytes (1.0 MB) copied, 0.0481317 s, 21.8 MB/s
>> [root@localhost btrfs]# sync
>> [root@localhost btrfs]# df -h /dev/sdb8
>> Filesystem            Size  Used Avail Use% Mounted on
>> /dev/sdb8              75G  1.1M   75G   1% /mnt/btrfs
>>
>> With raid-1, the used space may be double (about 2.2M), I think.
>>
>>
>>
>>> I think the partition size is no correct since this is a raid-1 partition.
>>> The size should be half.
>>> Is this a bug?
>>>
>>> -Shen
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>
>>>
>> --
>> Best Regards,
>> --------------------------------------------------
>> Shen Feng
>> Development Dept.I
>> Nanjing Fujitsu Nanda Software Tech. Co., Ltd.(FNST)
>> 8/F., Civil Defense Building, No.189 Guangzhou Road,
>> Nanjing, 210029, China
>> PHONE: +86-25-86630566-950
>> COINS: 79955-950
>> FAX: +86-25-83317685
>> MAIL: shen@cn.fujitsu.com
>> -------------------------------------------------
>> This communication is for use by the intended recipient(s) only and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not an intended recipient of this communication, you are hereby notified that any dissemination, distribution or copying hereof is strictly prohibited.  If you have received this communication in error, please notify me by reply e-mail, permanently delete this communication from your system, and destroy any hard copies you may have printed.
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
> 
> 
> 

-- 
Best Regards,
--------------------------------------------------
Shen Feng
Development Dept.I
Nanjing Fujitsu Nanda Software Tech. Co., Ltd.(FNST)
8/F., Civil Defense Building, No.189 Guangzhou Road,
Nanjing, 210029, China
PHONE: +86-25-86630566-950
COINS: 79955-950
FAX: +86-25-83317685
MAIL: shen@cn.fujitsu.com
-------------------------------------------------
This communication is for use by the intended recipient(s) only and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not an intended recipient of this communication, you are hereby notified that any dissemination, distribution or copying hereof is strictly prohibited.  If you have received this communication in error, please notify me by reply e-mail, permanently delete this communication from your system, and destroy any hard copies you may have printed.

Re: A bug with multiple devices?

[Chris Samuel]

[-- Attachment #1: Type: text/plain, Size: 905 bytes --]

On Wed, 31 Dec 2008 12:43:13 pm Shen Feng wrote:

> The problem still exists with -d option.

As Chris Mason wrote in response to my query on this it's a known issue that's 
a consequence of different chunks potentially having different RAID levels.

The RAID level itself does appear to work - I tested it by copying a kernel 
image onto it, unmounted it, used debug-tree to work out the offset for it on 
one device of the RAID and blew a section away with dd.  I remounted it and 
verified that the MD5 checksums still agreed with the original and there were 
kernel messages complaining about checksum mismatches showing that the damage 
had been detected and worked around.

cheers,
Chris
-- 
 Chris Samuel  :  http://www.csamuel.org/  :  Melbourne, VIC

This email may come with a PGP signature as a file. Do not panic.
For more info see: http://en.wikipedia.org/wiki/OpenPGP


[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 481 bytes --]