Quote:
Originally Posted by AlexFS
It's a technician responsibility to take care of the server, and come up with prevention in case shit hit's the fan as an account grows. Webmasters don't need to know jack-shit about servers as it's not their field of expertise. The downtime and no prevention, backup plan or proper solution is straight up bullshit.  |
Customer has backups, we can perform a full bare metal restore and restore their server to new equipment, that isn't the problem. The problem is the large amount of data being restored if we needed to would take.. many many hours to restore... So yes everything is in place for hardware failures etc, there is a hard choice to make and to figure out what will take less time to resolve, letting a server finish its simple check desk routine .. or restore it.. in this case.. letting it finish its check disk was the optimal solution..
And as I write this the customers server is back online and functioning just fine.
For those of you who don't know what fsck is, it is a utility short for file system check, it checks the drive for bad sectors, bad inodes, bad mappings to inodes etc. The slower the drives and greater amount of data you have the longer this process takes, its a 5 step process. In specific situations the linux operating system _forces_ you to do this. This is due to a file system being journalized.
Sure there are alternative file systems that do not have journalized file systems like ext3/4 suchs as riserFS or XFS. but frankly those are a bit inferior for day to day operation and have their own specific uses.