Hi,

Ok so we have this red hat machine at my work. It's been up since 2007 and never gave problems whatsoever. All updates have always been correctly installed even the latest ones.

attached file yum.txt

I don't know exactly when but arround october 29th we had this system freeze. We had to manually restart the machine.

attached file dmesg.txt

What's with the "ext3_orphan_cleanup: deleting unreferenced inode" message?

Am I loosing my disk or my filesystem? Should I prepare for the worst (I already backuped what needed to be saved).

Output from lscpi comand for hardware specs: attached lspci.txt

System kernel: 2.6.18-164.2.1.el5

Anyone can give some directions to understand if system is about to fail?

Attached Files

	yum.txt (1.1 KB, 16 views)
	dmesg.txt (48.7 KB, 12 views)
	lspci.txt (3.2 KB, 11 views)

Well, two years of constant uptime is a good bit. The inode messages could be from the system not being shut down properly, but the best place to look for suspicious things is /var/log/messages. Look for anything related to the hard drive device name.

Two years isn't very long, but the disk could be failing. You've taken the good steps of making sure you've got a backup, and making sure you're prepared for a crash. And that isn't an "if"...it's a "when".. All systems die eventually. If you see drive-related failures, I would try to schedule a time to replace the drives, before an unplanned outage. And if you're not short on $$$, I'd suggest upgrading the entire server, too...hardware is cheap these days, and you can migrate the services/files over at your leisure.

What are the best steps in order to understand if my disk is going to fail?

I have RAID 1 Arrayd (2 disks) with LVM configuration.

smartctl is very accurate

Your files do not suggest impending drive failure.

#man hdparm

use -I option

post output