Showing posts with label Defect. Show all posts
Showing posts with label Defect. Show all posts

2011-12-31

One bug in SED.EXE and another in MS-DOS cmd.exe DEL command

That was a fun day, a WTF one (What's The Fun?!). Wasted at least an hour on troubleshooting third party applications... -_-'

Damn'.


Sed.exe temporary file bug


The utility 'sed.exe' (from GnuWin32) helped me in modifying a file during a headless installation process (changing a property in a "web.config" file). Until there, it was fairly cool.


But here comes the less fun part, I noticed that the executable creates a temporary file right in the middle of its working directory. This is a bug listed and commented here:


From what it looks like, the tool may have regressed in Windows 7 again. Unless it's because of the 64 bit system and the WoW emulation...? Anyway.

The temporary file is not that random, first we know where it is located, second, its name is made of 3 known characters 'sed' and 6 random alphanumeric ones. In regular expressions it would be something like

/sed[a-zA-Z0-9]{6}/

So let's rock'n'roll with a workaround. Let's call:

$> DEL /q sed??????

Let's call it in a directory that contains a file sedE1, a file sedE1f8fd, and a file sedE1f8fdBIS. What happens. All three are deleted. WTF?!


Wildcards issue


After further investigations, "sed????" works fine. So what is wrong?

It seems that, as far as wildcards are concerned:

"?????????" <=> "*"

The explanation I found is that DEL works still on the 8.3 character file name model.

So finally the only workaround is to wrap sed.exe into a batch file:


2008-12-15

RAID failure, loss of data and anger

RAID is not a good alternative to backing up data. Data may become damaged or destroyed without harm to the drive(s) on which they are stored. For example, some of the data may be overwritten by a system malfunction; a file may be damaged or deleted by user error or malice and not noticed for days or weeks; and, of course, the entire array is at risk of physical damage.

Note that a RAID controller itself can become the single point of failure within a system.