Red Hat

It appears some vital programs have been corrupted. Do you have a backup?

no... I have no backup, this system base on mirror RAID.
May be I can copy some program from distributive DVD for this version?

Mirror RAID is not a backup. It's less sensitive to disk failure but, as you've seen, there's other ways for filesystems to go wrong. It's possible. We've got no idea how deep this corruption goes, though; it's possible a reinstall may be needed to fix this. Boot a livecd and get your vital data backed up before you try anything else.

When I performed boot from livecd and mount the /dev/sdaX partitions I have found that /sbin/mingetty work, and have not found any in lost+found

Now I can't fix problem with inoe 48912, I have tried to run e2fsk more and more from livecd for root /dev/sda1 (in ummount mode course), but error still exist...

Keep this up and it may be completely unable to mount soon, making recovery next to impossible. Stop MESSING with it for a few minutes and BACK IT UP!

Is this a hardware or software raid?

This is the hardware RAID - on Dell PowerEdge 2850 via LSI SCSI RAID adaptor.

A few weeks ago the server goes unresponsive, it is powered on but does not show to the display cannot be pinged so no remote access, on reboot server says memory/battery problems were detected, the adapter has recovred but cached datas was lost, hit any key to conitnue, reboot continues fine, linux and oracle run fine, and then few hours later server goes into same state, same message on reboot. Server has been online for just over 3 years.
I have tied to replace memory but the same result.

did you try change to raid controller wıth same model?