www.mamboteam.com
International Union of Elevator Constructors - Local 139 - Orlando, Florida Click for Orlando, Florida Forecast
Home arrow Articles arrow computers arrow How we salvaged the Web site from a failing HD
Thursday, 04 December 2008
 
 
Main Menu
Home
News
News Flashes
Articles
School
Forum and Guestbook
Links
Contact Us
Search
About this Site
Secure Login Form





Lost Password?
No account yet? Register
Newsflash
Please consider registering with the web site.  You need to be a registered user before we can UPGRADE your account to the IUEC Local 139 membership level.

Once your at this level, we can authorize and enable your account so that you can submit your own articles, images and photographs and this will also allow you access to the IUEC Local 139 private areas and forums that the general public and normal registered users are not allowed to view.
 
How we salvaged the Web site from a failing HD PDF Print E-mail
User Rating: / 0
PoorBest 
Written by Luther Pollok   
Sunday, 26 February 2006
Article Index
How we salvaged the Web site from a failing HD
DD_Rescue to the Rescue
My Game Plan
Running DD_Rescue
Checking the New Drive
Test Boot the New Drive
Summry of the Process
References

Once I had my copy on the new hard drive I ran fsck on each partition and the only one that failed was the partition that had the web site on it.

After some more reading about e2fsck I decided that I needed to use the -b option and point to the one of the backup journal entries on that partition.

Let me digress momentarily here to say that a EXT3 formatted partition is an EXT2 formatted partition with journal functionality added.  What this does for you is that whenever the operating system wirtes to the hard drive it also wirtes to the journal and records every transaction with with the hard drive, so that just n case of a situation similar to this the operating system can repair the file system in that partiton.  In a vague conceptual way, EXT3 is very similar to NTFS in things that is handled for the partition to protect data.

After running e2fsck -b nnnnnnn /dev/sdb7 (I forget the actual value I used now for nnnnnnn).  I could have probably used the 'yes for all' option as I did not know what the values should have been anyway and as it turned out I had to keep clicking "y" anyway.

When that was completed, I had a partition I could mount, inspect, copy and write to.  To my suprise, about 99.9% of all files were right where I left them.  Most of the problem appears to have occurred in the primary journal and some directory structures which did not take long to correct.



Last Updated ( Saturday, 10 June 2006 )
 
Next >
 
Top! Top!