Skip to contentSkip to conversation

BTRFS 3.8.7 Kernel Crash Report‏

DISCUSSION Mark Ridley 4 participants 14 messages Apr 22, 2013 Btrfs file system

lore
Share & tools
lore
#0 lore raw

Hi,

I have been having a kernel crash since kernel 3.3.2 and hoped it had been fixed by the time 3.8.7
(FC18) but no joy.

I take backups of large image files (around 100GB) and btrfs snapshot them.

If I then use rsync --inplace to update the images, I get a btrfs stack trace and btrfs hangs:

This happens every night.

Here is the stack trace:

Any idea what is causing this?

Thanks,

Mark

Apr 22 13:29:29 bkserver kernel: [66320.561671] ------------[ cut here ]----
--------
Apr 22 13:29:29 bkserver kernel: [66320.561818] kernel BUG at fs/btrfs/
ctree.c:2952!
Apr 22 13:29:29 bkserver kernel: [66320.561906] invalid opcode: 0000 
[#1] SMP
Apr 22 13:29:29 bkserver kernel: [66320.562212] Modules linked in:
 ppdev parport_pc vmxnet3

coretemp vmw_balloon parport shpchp i2c_piix4 microcode btrfs zlib_deflate libcrc32c
vmwgfx ttm
drm vmw_pvscsi i2c_core

Apr 22 13:29:29 bkserver kernel: [66320.562538] CPU 6Apr 22 
13:29:29 bkserver kernel: 
[66320.562605] Pid: 30965, comm: btrfs-endio-wri Not tainted 
3.8.7-201.fc18.x86_64 #1 VMware, 
Inc. VMware Virtual Platform/440BX Desktop Reference Platform
Apr 22 13:29:29 bkserver kernel: [66320.562820] RIP: 0010:[<ffffffffa00bd5c1>]  
[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]Apr 22 13:29:29
 bkserver kernel: 
[66320.563046] RSP: 0018:ffff88014763dae8  EFLAGS: 00010286
Apr 22 13:29:29 bkserver kernel: [66320.563147] RAX: 00000000ffffffff RBX:
 000000000000001a 
RCX: 0000000ce75ee000
Apr 22 13:29:29 bkserver kernel: [66320.563258] RDX: 00000000ffffffff RSI:
 ffff88014763dc16 RDI: 
ffff88014763dac7
Apr 22 13:29:29 bkserver kernel: [66320.563372] RBP: ffff88014763db48 R08: 
0000000000004b58 
R09: 00000ce75e80006c
Apr 22 13:29:29 bkserver kernel: [66320.563475] R10: 6c0000000000004b R11: 
0000000ce75e8000 R12: ffff88014763db07
Apr 22 13:29:29 bkserver kernel: [66320.563586] R13: ffff880101d151b0 R14: 
ffff88014763dc16 
R15: ffff8800afb69d90
Apr 22 13:29:29 bkserver kernel: [66320.563729] FS:  
0000000000000000(0000) 
GS:ffff8803bfd80000(0000) knlGS:0000000000000000
Apr 22 13:29:29 bkserver kernel: [66320.563842] CS:  0010 DS: 0000 ES: 0000 CR0: 
000000008005003b
Apr 22 13:29:29 bkserver kernel: [66320.563931] CR2: 0000000000440be0 
CR3: 
00000003ab6ab000 CR4: 00000000000007e0
Apr 22 13:29:29 bkserver kernel: [66320.564051] DR0: 0000000000000000
 DR1: 

0000000000000000 DR2: 0000000000000000

Apr 22 13:29:29 bkserver kernel: [66320.564185] DR3: 0000000000000000
 DR6: 00000000ffff0ff0 
DR7: 0000000000000400
Apr 22 13:29:29 bkserver kernel: [66320.564298] Process btrfs-endio-wri 
(pid: 30965, threadinfo 
ffff88014763c000, task ffff8803ab79aec0)
Apr 22 13:29:29 bkserver kernel: [66320.564455] Stack:
Apr 22 13:29:29 bkserver kernel: [66320.564558]  ffff88014763db48 
ffff8800066a9800 
ffff880131387000 5800000000000000
Apr 22 13:29:29 bkserver kernel: [66320.564662]  6c0000000000004b 
0000000ce75e8000 
ffff880100000000 ffff8800afb69d90
Apr 22 13:29:29 bkserver kernel: [66320.564755]  ffff880101d151b0 
0000000000000a69 
0000000000000000 0000000ce75d8000
Apr 22 13:29:29 bkserver kernel: [66320.564865] Call Trace:
Apr 22 13:29:29 bkserver kernel: [66320.564948]  [<ffffffffa00f1b50>] 
__btrfs_drop_extents+0x460/0xbc0 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.565048]  [<ffffffffa00cbe46>] ? 
run_clustered_refs+0x256/0xb50 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.565143]  [<ffffffffa00f740e>] ? 
alloc_extent_state+0x2e/0xc0 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.565300]  [<ffffffffa00f2d2b>] 
btrfs_drop_extents+0x6b/0xa0 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.565406]  [<ffffffffa00e31ab>] 
insert_reserved_file_extent.constprop.58+0x7b/0x2c0 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.566955]  [<ffffffffa00e06b5>] ? 
start_transaction+0x95/0x460 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.567792]  [<ffffffffa00e9004>] 
btrfs_finish_ordered_io+0x364/0x3f0 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.568534]  [<ffffffffa00e90a5>] 
finish_ordered_fn+0x15/0x20 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.569362]  [<ffffffffa0109e56>] 
worker_loop+0x136/0x580 
[btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.570101]  [<ffffffffa0109d20>] ? 
btrfs_queue_worker+0x300/0x300 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.570858]  [<ffffffff81081f40>]
 kthread+0xc0/0xd0
Apr 22 13:29:29 bkserver kernel: [66320.571561]  [<ffffffff81088f30>] ? 
set_security_override_from_ctx+0x50/0x50
Apr 22 13:29:29 bkserver kernel: [66320.572263]  [<ffffffff81010000>] ? 
perf_trace_xen_cpu_write_gdt_entry+0xa0/0x100
Apr 22 13:29:29 bkserver kernel: [66320.572924]  [<ffffffff81081e80>] ? 
kthread_create_on_node+0x120/0x120
Apr 22 13:29:29 bkserver kernel: [66320.573588]  [<ffffffff8165baac>] 
ret_from_fork+0x7c/0xb0
Apr 22 13:29:29 bkserver kernel: [66320.574225]  [<ffffffff81081e80>] ? 
kthread_create_on_node+0x120/0x120
Apr 22 13:29:29 bkserver kernel: [66320.574883] Code: 00 00 4c 89 e6 4c 

89 ff 48 98 48 8d 04 80
48 8d 54 80 65 e8 02 0c 04 00 4c 89 f6 4c 89 e7 e8 07 f2 ff ff 85 c0 0f 8f 5e ff ff ff <0f> 0b 0f 0b
66 66 2e 0f 1f 84 00 00 00 00 00 66 66 66 66 90 55

Apr 22 13:29:29 bkserver kernel: [66320.576813] RIP  [<ffffffffa00bd5c1>] 
btrfs_set_item_key_safe+0x141/0x150 [btrfs]
Apr 22 13:29:29 bkserver kernel: [66320.577476]  RSP <ffff88014763dae8>
Apr 22 13:29:29 bkserver kernel: [66320.579362] ---

[ end trace 0ae203d38f3d2f34 ]---

#1 lore raw
Quoted 6 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack trace and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and fsck should be able to at least detect it.

david

#2 lore raw

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing could be written to the volume.

Mark

Quoted 14 lines · David Sterba
Quoted 7 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and fsck should be able to at least detect it.

david

#3 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

Quoted 27 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing could be written to the volume.

Mark

Quoted 13 lines · David Sterba
Quoted 7 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and fsck should be able to at least detect it.

Quoted 6 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#4 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

No.

Without --repair pointed out some problems, but whats the point of knowing about the problems if they can't be fixed so I ran with --repair and it broke the volume.

Quoted 40 lines · Harald Glatt
Quoted 29 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 14 lines · David Sterba
Quoted 8 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and fsck should be able to at least detect it.

Quoted 7 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#5 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck output just from checking would be helpful in this case.

Quoted 49 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing about the problems if they can't be fixed so I ran with --repair and it broke the volume.

Quoted 40 lines · Harald Glatt
Quoted 29 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 14 lines · David Sterba
Quoted 8 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and fsck should be able to at least detect it.

Quoted 7 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#6 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

What does btrfs scrub do?

Is that meant to detect and fix problems?

Thanks,

Mark

Quoted 57 lines · Harald Glatt

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck output just from checking would be helpful in this case.

Quoted 52 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing
about the problems if they can't be fixed so I ran with --repair and it broke the volume.

Quoted 42 lines · Harald Glatt
Quoted 31 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight
corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 15 lines · David Sterba
Quoted 8 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and
fsck should be able to at least detect it.

Quoted 7 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#7 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

Only data errors (from CRC checks), maybe also some structure errors - I'm not sure. A remount should fix all errors. If it doesn't I think it's considered a bug - ultimately the goal is that no --repair should ever be required... Scrub is generally safe though, unlike btrfsck --repair.

Quoted 68 lines · Mark Ridley

What does btrfs scrub do?

Is that meant to detect and fix problems?

Thanks,

Mark

Quoted 57 lines · Harald Glatt

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck output just from checking would be helpful in this case.

Quoted 52 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing
about the problems if they can't be fixed so I ran with --repair and it broke the volume.

Quoted 42 lines · Harald Glatt
Quoted 31 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight
corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 15 lines · David Sterba
Quoted 8 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and
fsck should be able to at least detect it.

Quoted 7 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#8 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

Thanks for getting back to me.

Do you think remount fixes the keys reverted problem?

Quoted 79 lines · Harald Glatt

Only data errors (from CRC checks), maybe also some structure errors - I'm not sure. A remount should fix all errors. If it doesn't I think it's considered a bug - ultimately the goal is that no --repair should ever be required... Scrub is generally safe though, unlike btrfsck --repair.

Quoted 71 lines · Mark Ridley

What does btrfs scrub do?

Is that meant to detect and fix problems?

Thanks,

Mark

Quoted 60 lines · Harald Glatt

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck output just from checking would be helpful in this case.

Quoted 55 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing
about the problems if they can't be fixed so I ran with --repair and it
broke the volume.

Quoted 44 lines · Harald Glatt
Quoted 31 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight
corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 15 lines · David Sterba
Quoted 8 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and
fsck should be able to at least detect it.

Quoted 8 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#9 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

I think it won't... That would just be the goal eventually. If scrub sees no errors all the data should be in tact and your best bet to get things working perfectly again is to create a new filesystem and transfer the data into it.

I'm not a dev and I can't say if a btrfs-image for debugging purposes would be helpful in this case or not, maybe this kind of corrution has been fixed a long time ago and wouldn't happen again in up-to-date versions of btrfs anyway.

Quoted 86 lines · Mark Ridley

Thanks for getting back to me.

Do you think remount fixes the keys reverted problem?

Quoted 79 lines · Harald Glatt

Only data errors (from CRC checks), maybe also some structure errors - I'm not sure. A remount should fix all errors. If it doesn't I think it's considered a bug - ultimately the goal is that no --repair should ever be required... Scrub is generally safe though, unlike btrfsck --repair.

Quoted 71 lines · Mark Ridley

What does btrfs scrub do?

Is that meant to detect and fix problems?

Thanks,

Mark

Quoted 60 lines · Harald Glatt

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck output just from checking would be helpful in this case.

Quoted 55 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing
about the problems if they can't be fixed so I ran with --repair and it
broke the volume.

Quoted 44 lines · Harald Glatt
Quoted 31 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight
corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 15 lines · David Sterba
Quoted 8 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption and
fsck should be able to at least detect it.

Quoted 8 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#10 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

Any idea when the 3.9 kernel will be out.

I see there are lots of btrfs fixes going into it.

I'm currently running FC18 on 3.8.7, although 10 minutes ago it got updated to 3.8.8.

Quoted 101 lines · Harald Glatt

I think it won't... That would just be the goal eventually. If scrub sees no errors all the data should be in tact and your best bet to get things working perfectly again is to create a new filesystem and transfer the data into it.

I'm not a dev and I can't say if a btrfs-image for debugging purposes would be helpful in this case or not, maybe this kind of corrution has been fixed a long time ago and wouldn't happen again in up-to-date versions of btrfs anyway.

Quoted 89 lines · Mark Ridley

Thanks for getting back to me.

Do you think remount fixes the keys reverted problem?

Quoted 82 lines · Harald Glatt

Only data errors (from CRC checks), maybe also some structure errors - I'm not sure. A remount should fix all errors. If it doesn't I think it's considered a bug - ultimately the goal is that no --repair should ever be required... Scrub is generally safe though, unlike btrfsck --repair.

Quoted 74 lines · Mark Ridley

What does btrfs scrub do?

Is that meant to detect and fix problems?

Thanks,

Mark

Quoted 63 lines · Harald Glatt

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck output just from checking would be helpful in this case.

Quoted 57 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing
about the problems if they can't be fixed so I ran with --repair and it
broke the volume.

Quoted 46 lines · Harald Glatt
Quoted 33 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight
corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 17 lines · David Sterba
Quoted 9 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs
stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption
and
fsck should be able to at least detect it.

Quoted 8 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#11 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

Quoted 1 line · Mark Ridley

Any idea when the 3.9 kernel will be out.

Most likely next Sunday/Monday, unless something drastic shows up
this week (or there are too many fixes going in).

Quoted 4 lines

I see there are lots of btrfs fixes going into it.

I'm currently running FC18 on 3.8.7, although 10 minutes ago it got updated to 3.8.8.

Hugo.

Signature · 3 lines
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
  PGP key: 65E74AC0 from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
       --- We believe in free will because we have no choice. ---        
#12 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

In reply to Mark Ridley

I heard it's coming out next week.

Quoted 111 lines · Mark Ridley

Any idea when the 3.9 kernel will be out.

I see there are lots of btrfs fixes going into it.

I'm currently running FC18 on 3.8.7, although 10 minutes ago it got updated to 3.8.8.

Quoted 101 lines · Harald Glatt

I think it won't... That would just be the goal eventually. If scrub sees no errors all the data should be in tact and your best bet to get things working perfectly again is to create a new filesystem and transfer the data into it.

I'm not a dev and I can't say if a btrfs-image for debugging purposes would be helpful in this case or not, maybe this kind of corrution has been fixed a long time ago and wouldn't happen again in up-to-date versions of btrfs anyway.

Quoted 89 lines · Mark Ridley

Thanks for getting back to me.

Do you think remount fixes the keys reverted problem?

Quoted 82 lines · Harald Glatt

Only data errors (from CRC checks), maybe also some structure errors - I'm not sure. A remount should fix all errors. If it doesn't I think it's considered a bug - ultimately the goal is that no --repair should ever be required... Scrub is generally safe though, unlike btrfsck --repair.

Quoted 74 lines · Mark Ridley

What does btrfs scrub do?

Is that meant to detect and fix problems?

Thanks,

Mark

Quoted 63 lines · Harald Glatt

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck output just from checking would be helpful in this case.

Quoted 57 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing
about the problems if they can't be fixed so I ran with --repair and it
broke the volume.

Quoted 46 lines · Harald Glatt
Quoted 33 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight
corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so nothing
could be written to the volume.

Mark

Quoted 17 lines · David Sterba
Quoted 9 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs
stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

The check fails because it finds keys in reverted order. Given the conditions under which it happens I think it's an on-disk corruption
and
fsck should be able to at least detect it.

Quoted 8 lines
Quoted 1 line

david

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

This happened without --repair?

#13 lore raw

Re: BTRFS 3.8.7 Kernel Crash Report

For me anyway, I hope the fedora team don't take too long to make it available.

Thanks for all your help today.

Quoted 123 lines · Harald Glatt

I heard it's coming out next week.

Quoted 119 lines · Mark Ridley

Any idea when the 3.9 kernel will be out.

I see there are lots of btrfs fixes going into it.

I'm currently running FC18 on 3.8.7, although 10 minutes ago it got updated to 3.8.8.

Quoted 109 lines · Harald Glatt

I think it won't... That would just be the goal eventually. If scrub sees no errors all the data should be in tact and your best bet to get things working perfectly again is to create a new filesystem and transfer the data into it.

I'm not a dev and I can't say if a btrfs-image for debugging purposes would be helpful in this case or not, maybe this kind of corrution has been fixed a long time ago and wouldn't happen again in up-to-date versions of btrfs anyway.

Quoted 97 lines · Mark Ridley

Thanks for getting back to me.

Do you think remount fixes the keys reverted problem?

Quoted 90 lines · Harald Glatt

Only data errors (from CRC checks), maybe also some structure errors - I'm not sure. A remount should fix all errors. If it doesn't I think it's considered a bug - ultimately the goal is that no --repair should ever be required... Scrub is generally safe though, unlike btrfsck --repair.

Quoted 81 lines · Mark Ridley

What does btrfs scrub do?

Is that meant to detect and fix problems?

Thanks,

Mark

Quoted 70 lines · Harald Glatt

Yeah, --repair is not recommended as of now. Maybe posting the btrfsck
output just from checking would be helpful in this case.

Quoted 63 lines · Mark Ridley

No.

Without --repair pointed out some problems, but whats the point of knowing
about the problems if they can't be fixed so I ran with --repair and
it
broke the volume.

Quoted 51 lines · Harald Glatt
Quoted 39 lines · Mark Ridley

Thanks, David.

What causes this corruption and how can I fix it?

I'm very worried about running btrfs.fsck as last time it made a slight
corruption like this worse and the whole volume had to be trashed.

After fsck the "available space" on df ended up being negative so
nothing
could be written to the volume.

Mark

Quoted 12 lines · David Sterba
Quoted 10 lines · Mark Ridley

If I then use rsync --inplace to update the images, I get a btrfs
stack
trace
and btrfs hangs:

This happens every night.

[<ffffffffa00bd5c1>] btrfs_set_item_key_safe+0x141/0x150 [btrfs]

Quoted 8 lines

The check fails because it finds keys in reverted order. Given the
conditions under which it happens I think it's an on-disk corruption
and
fsck should be able to at least detect it.

david

Quoted 7 lines

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to [email protected] More majordomo info at
http://vger.kernel.org/majordomo-info.html

This happened without --repair?