headers
Andreas Dilger | 2 Jul 2008 09:28
Picon

Re: debugfs question

On Jun 28, 2008  23:50 +0300, Yamin, Yossi wrote:
> I am trying to read a file directly from the disk using debugfs utility.
> 
> I am running "stat" on the file I want, filter out IND and Bind blocks,
> and then copy the data blocks using dd directly from the Disk.
> 
> On small files it work'd (13MB).
> 
> On big files (3.5 , 440 GB)  the size is the same but md5sum get differ.
> 
> What am I doing wrong?

What you are doing wrong is that you aren't using the debugfs "dump"
command, which will do what you are trying to do.

Cheers, Andreas
--
Andreas Dilger
Sr. Staff Engineer, Lustre Group
Sun Microsystems of Canada, Inc.
Permalink | Reply | 1 comment |
headers
Jure Pečar | 2 Jul 2008 13:18

max fs size with 1k blocks


Hello,

what is max ext3 filesystem size when mke2fs is called with -i 1024 -b 1024?

If rhel5 supports 8T ext3, I assume that is with 4k blocks, right? So it is right to assume this number to be 2T
with 1k blocks?

What about rhel4?

Thanks for answers.

--

-- 

Jure Pečar
http://jure.pecar.org
Permalink | Reply | 1 comment |
headers
Yamin, Yossi | 3 Jul 2008 16:07

RE: debugfs question

Thanks Andreas,

I have a problem that for the file data I need the fs MetaData is corrupted.

I hope that because I am using same pre allocate logic all over my FS then I can get the Maping from other clear state FS file

And read the data directly from the corrupted FS.

Is that making sense?  

 

Thanks,

Yossi

 

-----Original Message-----
From: Andreas.Dilger <at> sun.com [mailto:Andreas.Dilger <at> sun.com] On Behalf Of Andreas Dilger
Sent: Wednesday, July 02, 2008 10:29 AM
To: Yamin, Yossi
Cc: ext3-users <at> redhat.com; tytso <at> mit.edu
Subject: Re: debugfs question

 

On Jun 28, 2008  23:50 +0300, Yamin, Yossi wrote:

> I am trying to read a file directly from the disk using debugfs utility.

>

> I am running "stat" on the file I want, filter out IND and Bind blocks,

> and then copy the data blocks using dd directly from the Disk.

>

> On small files it work'd (13MB).

>

> On big files (3.5 , 440 GB)  the size is the same but md5sum get differ.

>

> What am I doing wrong?

 

What you are doing wrong is that you aren't using the debugfs "dump"

command, which will do what you are trying to do.

 

 

Cheers, Andreas

--

Andreas Dilger

Sr. Staff Engineer, Lustre Group

Sun Microsystems of Canada, Inc

 

.

 

_______________________________________________
Ext3-users mailing list
Ext3-users <at> redhat.com
https://www.redhat.com/mailman/listinfo/ext3-users
Permalink | Reply | 0 comments |
headers
Eric Sandeen | 3 Jul 2008 19:36
Picon
Favicon
Gravatar

Re: max fs size with 1k blocks

Jure Pečar wrote:
> Hello,
> 
> what is max ext3 filesystem size when mke2fs is called with -i 1024 -b 1024?

32-bit block number container, so 2^32 * 1024 = 4T.
> 
> If rhel5 supports 8T ext3, 

16T actually, in 2.6.19 and beyond (or distros which backported the fixes)

> I assume that is with 4k blocks, right? 

Yep.

> So it is right to assume this number to be 2T with 1k blocks?

s/b 4T, see above.

> What about rhel4?

on pre-2.6.18/19 kernels there were really only 31 safe bits, so cut it
in half.

-Eric
Permalink | Reply | 0 comments |
headers
Pau Garcia i Quiles | 4 Jul 2008 18:16
Favicon

Nanosecond date resolution

Hello,

By default ext3 stores file with 1 second date resolution but I have  
read it is possible to enable big inode and that will improve  
resolution to 1 nanosecond.

How can I enable the 1-nanosecond date resolution? I've found some  
mails by Alex Tomas, Andi Kleen and Andreas Gruenbacher to the linux  
filesystems mailing list but no clear way to enable this feature. Will  
it harm performance? (I do know it consumes more space)

Thank you

--

-- 
Pau Garcia i Quiles
http://www.elpauer.org
(Due to my workload, I may need 10 days to answer)
Permalink | Reply | 1 comment |
headers
Eric Sandeen | 4 Jul 2008 19:04
Picon
Favicon
Gravatar

Re: Nanosecond date resolution

Pau Garcia i Quiles wrote:
> Hello,
> 
> By default ext3 stores file with 1 second date resolution but I have  
> read it is possible to enable big inode and that will improve  
> resolution to 1 nanosecond.
> 
> How can I enable the 1-nanosecond date resolution? I've found some  
> mails by Alex Tomas, Andi Kleen and Andreas Gruenbacher to the linux  
> filesystems mailing list but no clear way to enable this feature. Will  
> it harm performance? (I do know it consumes more space)
> 
> Thank you
> 

ext4 makes use of larger inodes for this purpose, but ext3 does not.

-Eric
Permalink | Reply | 0 comments |
headers
Iskandar Prins | 13 Jul 2008 01:01
Picon

Ext3 problem

Hey All,

I have a serious and weird problem with my harddrive and ext3. About a week ago I re-arranged my entire directory structure. I moved, deleted, renamed files and dirs. I did some reboots and shutdowns in the mean time. And all was fine, up until now.

At the moment I'm looking at the filesystem (onto which I made those changed last week) , but it seems I'm looking at the filesystem from a month ago ( june 20th).  

Now I have a couple of questions:

- How could this have happened?

- Is this a journaling problem? If so, how can I fix this and get the filesystem back from july 12th instead of looking at the filesystem how it was on june 20th.

I did a fsck/e2fsck on the umounted devs and they can't get any cleaner.

Regards,
Iskandar

_______________________________________________
Ext3-users mailing list
Ext3-users <at> redhat.com
https://www.redhat.com/mailman/listinfo/ext3-users
Permalink | Reply | 6 comments |
headers
Eric Sandeen | 13 Jul 2008 01:30
Picon
Favicon
Gravatar

Re: Ext3 problem

Iskandar Prins wrote:
> Hey All,
> 
> I have a serious and weird problem with my harddrive and ext3. About a
> week ago I re-arranged my entire directory structure. I moved, deleted,
> renamed files and dirs. I did some reboots and shutdowns in the mean
> time. And all was fine, up until now.
> 
> At the moment I'm looking at the filesystem (onto which I made those
> changed last week) , but it seems I'm looking at the filesystem from a
> month ago ( june 20th).  
> 
> Now I have a couple of questions:
> 
> - How could this have happened?

No idea.

> - Is this a journaling problem? 

It is, but...

> If so, how can I fix this and get the
> filesystem back from july 12th instead of looking at the filesystem how
> it was on june 20th.

... that's not what journaling does I'm afraid.

> I did a fsck/e2fsck on the umounted devs and they can't get any cleaner.

I don't suppose you changed hard drives and you're looking at the old
one? :)

-Eric
Permalink | Reply | 0 comments |
headers
Levent Serinol | 13 Jul 2008 11:14
Picon

Re: indexing symbolic links

you can use inotify and register it to notify you when a symbolic link
created or unlinked. By this way you can put or remove names from your
database automatically, with a small C,perl,etc. program :)

2008/6/22 Mag Gam <magawake <at> gmail.com>:
> Unfortunately, tracking space wasn't me goal. I want to keep track of my
> symbolic links :-)
>
>
>
> On Sun, Jun 22, 2008 at 3:04 PM, Alex Bligh <alex <at> alex.org.uk> wrote:
>>
>>
>> --On 22 June 2008 09:12:26 -0400 Mag Gam <magawake <at> gmail.com> wrote:
>>
>>> At my university, we have physical storage in a filesystem, and we assign
>>> professors and students space by doing a symbolic link. Basically I want
>>> to keep track of physical storage with virtual/logical storage. Thats why
>>> I ask :-)
>>
>> If you want to track space usage, I suggest you track it using quota
>> or similar. "man quota" will give you a start.
>>
>> Alex
>
>
> _______________________________________________
> Ext3-users mailing list
> Ext3-users <at> redhat.com
> https://www.redhat.com/mailman/listinfo/ext3-users
>
Permalink | Reply | 1 comment |
headers
Mag Gam | 13 Jul 2008 18:04
Picon

Re: indexing symbolic links

OH! Very nice solution. I guess I can listen to that. Thats a cool way
of doing it.

On Sun, Jul 13, 2008 at 5:14 AM, Levent Serinol <lserinol <at> gmail.com> wrote:
> you can use inotify and register it to notify you when a symbolic link
> created or unlinked. By this way you can put or remove names from your
> database automatically, with a small C,perl,etc. program :)
>
> 2008/6/22 Mag Gam <magawake <at> gmail.com>:
>> Unfortunately, tracking space wasn't me goal. I want to keep track of my
>> symbolic links :-)
>>
>>
>>
>> On Sun, Jun 22, 2008 at 3:04 PM, Alex Bligh <alex <at> alex.org.uk> wrote:
>>>
>>>
>>> --On 22 June 2008 09:12:26 -0400 Mag Gam <magawake <at> gmail.com> wrote:
>>>
>>>> At my university, we have physical storage in a filesystem, and we assign
>>>> professors and students space by doing a symbolic link. Basically I want
>>>> to keep track of physical storage with virtual/logical storage. Thats why
>>>> I ask :-)
>>>
>>> If you want to track space usage, I suggest you track it using quota
>>> or similar. "man quota" will give you a start.
>>>
>>> Alex
>>
>>
>> _______________________________________________
>> Ext3-users mailing list
>> Ext3-users <at> redhat.com
>> https://www.redhat.com/mailman/listinfo/ext3-users
>>
>
Permalink | Reply | 0 comments |

Gmane