Hello.

I have a Fedora Core 3 File server with a RAID 5 configuration using a RocketRAID card.

A couple of times it has gone "read-only" and I needed to run fsck to restore the file system. This was on my production server so I lost a lot of man hours in downtime.

My question is if this is caused by a physical problem in the hard drives or is it a logical problem with how the file system writes the data to the disk? Does this happen at the sector level or data writing level?

Thank you in advance. Any help will be appreciated.

atrejo.

It probably ran into an error and was mounted with "errors=remount-ro"

Look into failing disks

If there were any failing disks, wouldn't the array raid 5 algorithm or system kicked and kept on working with error messages about a failing disk?

Thank you.

Dear
What is your log files. /var/log/messages.
Also check output of 'dmesg' command

Thanks