All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ryusuke Konishi <ryusuke-sG5X7nlA6pw@public.gmane.org>
To: liste-ZXOCcuNttmoBe96aLqz0jA@public.gmane.org
Cc: linux-nilfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: Bug : reuse same inode
Date: Tue, 08 Feb 2011 20:42:30 +0900 (JST)	[thread overview]
Message-ID: <20110208.204230.73657026.ryusuke@osrg.net> (raw)
In-Reply-To: <20110207204232.364e4c5f@micro24>

Hi,

Thank you for the detail report!  It's really helpful.

I understood the zero link count inode was causing a chain reaction.

And the inode in question at cp15731 also seems to be overridden
because gid, size, and timestamp all differs from that file at
cp15730.

> cp15730
> -------
> mnt/snap5/cut:
> total 4
> 14 -rw-r--r-- 1 philippe 1000 47 25 janv. 22:33 f5_20110125_202053.ts(0).Xcl

> cp15731
> -------
> mnt/snap7/cut:
> total 9223372036848683032
> 14 -rw-r--r-- 0 philippe users 6214528000 24 janv. 01:30 f5_20110125_202053.ts(0).Xcl

Can't you find another file with the inode number 14 in the snapshot
cp15731 or cp15730 ?

I reviewed back unlink() implementation, but haven't found any issues
so far.  The problem may be rooted much deeper.

'stat' command may give us much information than 'ls -l' in this case.


Thanks,
Ryusuke Konishi


On Mon, 7 Feb 2011 20:42:32 +0100, Philippe CHATAIGNON wrote:
> Le Mon, 07 Feb 2011 13:02:14 +0900 (JST),
> Ryusuke Konishi <ryusuke-sG5X7nlA6pw@public.gmane.org> a écrit :
> 
> > Can you track the past state of this file from earlier checkpoints ?
> Just before recording :
> philippe@micro11:~$ sudo mount -t nilfs2 -r -o cp=16940 /dev/mapper/vg11a-video mnt/snap5
> philippe@micro11:~$ ls -li mnt/snap5/**
> 18 -rw-r--r-- 1 philippe users 4434  9 janv. 15:45 mnt/snap5/channels.conf
> 
> mnt/snap5/capture:
> total 19452056
> 16 -rw-r--r-- 1 philippe users 8606489600  9 janv. 15:42 f3_20110107_205154.ts
> 15 -rw-r--r-- 1 philippe users 4688870400 19 janv. 23:14 f4_20110119_204007.ts
> 17 -rw-r--r-- 1 philippe users 6545408000  9 janv. 15:45 tf1hd_20101219_203305_schrek2.ts
> 
> mnt/snap5/cut:
> total 4
> 19 -rw-r--r-- 1 philippe  1000        156 29 janv. 21:16 canal_20110129_181951.ts(0).Xcl
> 14 -rw-r--r-- 0 philippe users 4838067200 30 janv. 00:58 f5_20110125_202053.ts(0).Xcl
> 
> Start of recording :
> philippe@micro11:~$ sudo mount -t nilfs2 -r -o cp=16941 /dev/mapper/vg11a-video mnt/snap6
> philippe@micro11:~$ ls -li mnt/snap6/**                                                  
> 18 -rw-r--r-- 1 philippe users 4434  9 janv. 15:45 mnt/snap6/channels.conf
> 
> mnt/snap6/capture:
> total 19454420
> 14 -rw-r--r-- 1 philippe users    2406400 30 janv. 17:45 f2_20110130_174456.ts
> 16 -rw-r--r-- 1 philippe users 8606489600  9 janv. 15:42 f3_20110107_205154.ts
> 15 -rw-r--r-- 1 philippe users 4688870400 19 janv. 23:14 f4_20110119_204007.ts
> 17 -rw-r--r-- 1 philippe users 6545408000  9 janv. 15:45 tf1hd_20101219_203305_schrek2.ts
> 
> mnt/snap6/cut:
> total 2368
> 19 -rw-r--r-- 1 philippe  1000     156 29 janv. 21:16 canal_20110129_181951.ts(0).Xcl
> 14 -rw-r--r-- 1 philippe users 2406400 30 janv. 17:45 f5_20110125_202053.ts(0).Xcl
> 
> That's the problem : inode 14 is reused because there was a file with 0 link (f5_20110125_202053.ts(0).Xcl) 
> in another directory. So I tracked earlier and found that :
> 
> cp15730
> -------
> philippe@micro11:~$ ls -li mnt/snap5/**
> 18 -rw-r--r-- 1 philippe users 4434  9 janv. 15:45 mnt/snap5/channels.conf
> 
> mnt/snap5/capture:
> total 19452056
> 16 -rw-r--r-- 1 philippe users 8606489600  9 janv. 15:42 f3_20110107_205154.ts
> 15 -rw-r--r-- 1 philippe users 4688870400 19 janv. 23:14 f4_20110119_204007.ts
> 17 -rw-r--r-- 1 philippe users 6545408000  9 janv. 15:45 tf1hd_20101219_203305_schrek2.ts
> 
> mnt/snap5/cut:
> total 4
> 14 -rw-r--r-- 1 philippe 1000 47 25 janv. 22:33 f5_20110125_202053.ts(0).Xcl
> 
> All is OK
> 
> cp15731
> -------
> philippe@micro11:~$ ls -li mnt/snap7/**
> 18 -rw-r--r-- 1 philippe users 4434  9 janv. 15:45 mnt/snap7/channels.conf
> 
> mnt/snap7/capture:
> total 19452056
> 16 -rw-r--r-- 1 philippe users 8606489600  9 janv. 15:42 f3_20110107_205154.ts
> 15 -rw-r--r-- 1 philippe users 4688870400 19 janv. 23:14 f4_20110119_204007.ts
> 17 -rw-r--r-- 1 philippe users 6545408000  9 janv. 15:45 tf1hd_20101219_203305_schrek2.ts
> 
> mnt/snap7/cut:
> total 9223372036848683032
> 14 -rw-r--r-- 0 philippe users 6214528000 24 janv. 01:30 f5_20110125_202053.ts(0).Xcl
> 
> Size of file f5_20110125_202053.ts(0).Xcl is incorrect and it appears with 0 link.
> 
> 
> cp15732
> -------
> philippe@micro11:~$ ls -li mnt/snap2/**
> 18 -rw-r--r-- 1 philippe users 4434  9 janv. 15:45 mnt/snap2/channels.conf
> 
> mnt/snap2/capture:
> total 19452056
> 14 -rw-r--r-- 1 philippe users          0 29 janv. 18:19 canal_20110129_181951.ts
> 16 -rw-r--r-- 1 philippe users 8606489600  9 janv. 15:42 f3_20110107_205154.ts
> 15 -rw-r--r-- 1 philippe users 4688870400 19 janv. 23:14 f4_20110119_204007.ts
> 17 -rw-r--r-- 1 philippe users 6545408000  9 janv. 15:45 tf1hd_20101219_203305_schrek2.ts
> 
> mnt/snap2/cut:
> total 0
> 14 -rw-r--r-- 1 philippe users 0 29 janv. 18:19 f5_20110125_202053.ts(0).Xcl
> 
> 
> Inode 14 is used by 2 files. When I deleted the file canal_20110129_181951.ts, we have
> the situation of the cp 16940 (the first of this mail) and a file with 0 link.
> 
> Problem is in cp15731, just before the creation of file canal_20110129_181951.ts.
> 
> PhC
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

  reply	other threads:[~2011-02-08 11:42 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-02-06 19:09 Bug : reuse same inode Philippe CHATAIGNON
2011-02-07  4:02 ` Ryusuke Konishi
     [not found]   ` <20110207.130214.221723444.ryusuke-sG5X7nlA6pw@public.gmane.org>
2011-02-07 19:42     ` Philippe CHATAIGNON
2011-02-08 11:42       ` Ryusuke Konishi [this message]
     [not found]         ` <20110208.204230.73657026.ryusuke-sG5X7nlA6pw@public.gmane.org>
2011-02-08 20:32           ` Philippe CHATAIGNON
2011-02-10  5:01             ` Ryusuke Konishi
     [not found]               ` <20110210.140105.33383103.ryusuke-sG5X7nlA6pw@public.gmane.org>
2011-02-10 22:32                 ` Philippe CHATAIGNON
2011-02-18  7:21                   ` Ryusuke Konishi
     [not found]                     ` <20110218.162145.57461053.ryusuke-sG5X7nlA6pw@public.gmane.org>
2011-02-18 19:57                       ` Philippe CHATAIGNON
2011-02-19 13:22                         ` Ryusuke Konishi
     [not found]                           ` <20110219.222232.05855196.ryusuke-sG5X7nlA6pw@public.gmane.org>
2011-02-19 15:26                             ` Philippe CHATAIGNON

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20110208.204230.73657026.ryusuke@osrg.net \
    --to=ryusuke-sg5x7nla6pw@public.gmane.org \
    --cc=linux-nilfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=liste-ZXOCcuNttmoBe96aLqz0jA@public.gmane.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.