search

danpedron / snake-os

Automatically exported from code.google.com/p/snake-os
0 stars 1 forks source link

Disk remount as read only #195

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
What steps will reproduce the problem?
1. Wait until disk goes to sleep
2. Woke it up
3. Try to wright on it, or open a file

What is the expected output? What do you see instead?

In an expected output I would be able to write to my disk, instead it says that 
my disk is read-only. In snake log I get something like this:
Aug 28 22:22:40 kernel: Remounting filesystem read-only
Aug 28 22:22:40 kernel: EXT3-fs error (device sdb1): ext3_journal_start_sb: 
Detected aborted journal
Aug 28 22:22:40 kernel: ext3_abort called.

What version of the product are you using? On what operating system?

Kernel version: 2.6.16-gazineu
NAS version:    SNAKE OS V1.3.2 (20101130)

Please provide any additional information below.

My disk : Seagate 2TB External Desktop Hard Drive (ST320005EXA101-RK)
My NAS :  WLX-652 http://www.dealextreme.com/details.dx/sku.20383

I've done many e2fsck on my disk and it doesn't resolve the problem, neither 
did a journal nuke.

My disk is brand new, I’ve bought it less than a month ago and I’ve not 
drop it or anything like that.

Original issue reported on code.google.com by Dion.Kev...@gmail.com on 1 Sep 2011 at 2:19

GoogleCodeExporter commented 9 years ago 
Is there anything before those messages in log? 

If it works before the drive goes to sleep you can maybe disable power 
management/sleep with the attached hdparm. I'm not sure if this works for 
external drives though..

Original comment by stefansc...@googlemail.com on 1 Sep 2011 at 9:05

Attachments:

GoogleCodeExporter commented 9 years ago 
More log:
Aug 28 22:22:40 kernel: Remounting filesystem read-only
Aug 28 22:22:40 kernel: EXT3-fs error (device sdb1): ext3_journal_start_sb: 
Detected aborted journal
Aug 28 22:22:40 kernel: ext3_abort called.
Aug 28 22:21:06 kernel: Aborting journal on device sdb1.
Aug 28 22:21:06 kernel: EXT3-fs error (device sdb1): ext3_find_entry: reading 
directory #1761295 offset 0
Aug 28 22:21:06 kernel: end_request: I/O error, dev sdb, sector 759839087
Aug 28 22:21:06 kernel: sd 1:0:0:0: Device not ready.

I don't know if I want my drive to be running forever, I bet it is not good for 
life time durability...

I've come across something like using "tune2fs -e continue /dev/(device)" to 
make the disk continue instead of remounting read-only when it encounter an 
error... But I don't know if that is secure.

Original comment by Dion.Kev...@gmail.com on 2 Sep 2011 at 4:03

GoogleCodeExporter commented 9 years ago 
Yeah.. disabling sleep isn't really optimal. 

It sounds like the the drive can't be woken up after it goes to sleep. The 
filesystem is probably ok it just can't be accessed anymore.

Your problem sounds suspicously like this one:
https://bugzilla.redhat.com/show_bug.cgi?id=433287

If so, it would require a small kernel patch to fix.

Original comment by stefansc...@googlemail.com on 2 Sep 2011 at 8:51

GoogleCodeExporter commented 9 years ago 
Yeah it seems like my problem looks like that one. Except that I can acces my 
file when the drive is remounted in read-only. 

I'm a little lost in all the kernel update, is it possible to do it for snake 
os?

Original comment by Dion.Kev...@gmail.com on 3 Sep 2011 at 2:35

GoogleCodeExporter commented 9 years ago 
The kernel can be flashed separately but it's probably safer to do it with a 
firmware upgrade.

I can put up a snapshot build but can only verify that it works on my own box 
(a ritmo ln-86bt). So there's a risk that it could brick other ones.

Also, transmission is in svn at the moment. If you use that it wouldn't be much 
of an upgrade.

Original comment by stefansc...@googlemail.com on 5 Sep 2011 at 11:58

GoogleCodeExporter commented 9 years ago 
I don't use transmission at all, but I'm not sure I want to risk my NAS.

Will the next official version of snake os include an up to date kernel?

I saw that it's suppose to be fix in 2.6.24-rc4 Kernel... As now Snake os 
2.6.16-gazineu

Original comment by Dion.Kev...@gmail.com on 6 Sep 2011 at 1:18

GoogleCodeExporter commented 9 years ago 
Oups .. I meant transmission is _missing_ in svn

The next version will probably stick to 2.6.16. But from what I can tell the 
fix should work there too. 

I'm not sure if it will even help though. The bug report mentions that he could 
wake up the disk with fdisk. Does that work for you?

Original comment by stefansc...@googlemail.com on 6 Sep 2011 at 8:54

GoogleCodeExporter commented 9 years ago 
When I run fdisk to wake up the disk I get this log : 

Sep 6 15:27:59 kernel: Buffer I/O error on device sdb, logical block 0
Sep 6 15:27:59 kernel: end_request: I/O error, dev sdb, sector 0
Sep 6 15:27:59 kernel: sd 1:0:0:0: Device not ready.

But if I try to rename a file or add one, the disk is not remount as Read-Only.

Original comment by Dion.Kev...@gmail.com on 6 Sep 2011 at 7:35

GoogleCodeExporter commented 9 years ago 
If anyone encounters a problem like the one above, it has been fixed in the 
latest firmware snakeos-V1.3.2-20111019. I've just upgraded to it and it works 
just fine waking up the disk after sleep.

Original comment by Dion.Kev...@gmail.com on 28 Dec 2011 at 7:46

GoogleCodeExporter commented 9 years ago 
Sweet.. so the fix really works.

Original comment by stefansc...@googlemail.com on 30 Dec 2011 at 3:42

GoogleCodeExporter commented 9 years ago 
Yup my disk now work as if I had disabled the sleep timer. Without the disk 
running 24\7.

Original comment by Dion.Kev...@gmail.com on 30 Dec 2011 at 10:36