How on earth do files seem to just disappear?

[glennzo]

Like the title says. How on earth do files seem to just disappear?

I have an NFS exported folder that contains several scripts. It is exported to a few different local IP addresses so that no matter what computer I'm using I can access the exported folder and it's contents. It is also mounted automatically in most cases. Now, all this has been working extremely well for several years but today, there are no files in this folder.

The most likely answer is that I unknowinglly deleted them but I just wonder if files in an exported folder have been known to disappear for no obvious reason. Poof!

Thank goodness for backups.

[Dan]

It's Electromagnetic Karma divergence caused by parasitic frequency oscillations. Can't be helped in this day and age of cellphones, wireless, microwave ovens, and government saturation of the populace by Q-bank torminus radions.

But, the good news is ... for the paltry price of only $80.00USD, I can help you solve your problem with a holographically magno-brushed meteoric nickel-iron impregnated computer cable interference-resistor clamp.

Wanna buy a catalog? <....>


On a more serious note ... impending drive failure?

[glennzo]

Originally Posted by Dan

On a more serious note ... impending drive failure?

I suppose that's a possibility.

[root@server glenn>$ smartctl -A /dev/sde
smartctl 5.43 2012-06-30 r3573 [i686-linux-3.4.4-4.fc16.i686.PAE] (local build)
Copyright (C) 2002-12 by Bruce Allen, http://smartmontools.sourceforge.net

=== START OF READ SMART DATA SECTION ===
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   058   052   006    Pre-fail  Always       -       42295055
  3 Spin_Up_Time            0x0003   096   096   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       518
  5 Reallocated_Sector_Ct   0x0033   100   100   036    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x000f   078   060   030    Pre-fail  Always       -       68381134
  9 Power_On_Hours          0x0032   058   058   000    Old_age   Always       -       36921
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   098   098   020    Old_age   Always       -       2179
194 Temperature_Celsius     0x0022   049   062   000    Old_age   Always       -       49
195 Hardware_ECC_Recovered  0x001a   058   052   000    Old_age   Always       -       42295055
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   182   000    Old_age   Always       -       271
200 Multi_Zone_Error_Rate   0x0000   100   253   000    Old_age   Offline      -       0
202 Data_Address_Mark_Errs  0x0032   100   253   000    Old_age   Always       -       0

Looks good to the untrained eye, especially the "WHEN FAILED" column.

[Chilly Willy]

Is it possible that a power glitch could have taken place? With all the storms (in MY area anyway) that is the first thing that came to mind.

Or, possibly pointer errors occurring in whatever you're using for the transfer routine?

[flyingfsck]

Howdy,

If you dual boot with Windows and let it suspend instead of shut down, then run Linux and change something on the NTFS disk, then awaken Windows, it will 'repair' the file system for you...

So, the moral of the story is, don't use suspend features if you dual boot. Even better, do not dual boot, use Virtualbox for Windows on Linux and let Windows run in a window, where it belongs.

[glennzo]

No dual booting of Windows on the server, no transfer routine since it's an NFS share and no power issues in the area yesterday, although there have been some severe thunderstorms in the past week.

[beaker_]

Just a guess but revisit your cron tasks and any other automated scripts you've edited this week. If you're paroniod or in doubt then; assuming its happening over the network, let wireshark log (filter for nfs) and dump it to a file; assuming nothing then try and write something with inotify and maybe catch it in your logs using time & date as a reference point... maybe?

[sea]

What i experienced recently was that i had expected my scripts to upload files to my NAS.
As that was 'run' by cron, i had expected them to be on the NAS, but they werent.
After the shockmoment, and unmount the NAS, i had a look at my local path structure, to figure they were properly copied to the "mounted" path, just that the path wasnt mounted.

At the time the NAS was mounted, the files in that folder got hidden, as they were 'overwritten' with the 'empty' content of the NAS, while they were there but localy.

[flyingfsck]

At the time the NAS was mounted, the files in that folder got hidden

Yup, done that before. I have taken to creating a file called 'NOT MOUNTED!' in every mount point directory.