when we mount CD on the unix, at that time directory that is used to mount must alredy be created ? Now everytime we want to use Cd, we can insert CD in the system and use mount point to access this CD ?
Can we just copy the content of the CD in another directory ? This directory has to be separate than mount point or can we copy content in the same mount point ?
Does the directory where content is being copied has to be already existing or can it be craeted at the time of copying ? Does CD need to be in CD drive in order to mount it ? Or I can create a mount and later on insert the CD in CD drive and read it from mount point ?
Do I have to mount everytime, I want to use CD ? Once I take the CD out, then if I want to use CD again, do I need to mount it again ?
what is meant by NFS mounting point without caching ? and NFS mounting point with caching ?
First of all, yes, a directory has to be pre-existing to be used as a mount point.
Secondly, you wouldn't want to copy a CD every time you used it. CD and DVD devices are rather slow.
Thirdly, there is an automount facility available in most Linux distros. Basically, a daemon scans the devices, detects media insertion and performs the mount command for you. On my system, it mounts on /media/volumename, where "volumename" is the volume name of the CD. If it has none, it mounts on /media/cdrom.
I think that these days, all that's part of the hotplug subsystem, which also handles things like USB "thumb" drives and wireless devices. Although I could be wrong and Your Mileage May Vary.
You cannot remove a device or media without unmounting it. Even if you physically force it (paper clip), the computer will "think" the device/media is still there. This is especially important when considering things like USB thumb devices and floppy disks. Part of the unmount process was to ensure that all data in the OS buffers actually got written to the device/media while it still was attached and able to receive that data. Unmounting also ensures that all files open on the media/device are closed (actually, it usually tests and fails rather than closing them itself).
Caching in NFS is like caching anywhere else. You keep copies of stuff nearby in fast-access storage to reduce the overhead involved in fetching from the "real", but slower actual storage. So an NFS cache keeps local copies of data that comes from network shares and ensures that (eventually), any changes to those local copies gets pushed back to the network shares. This is all pretty much transparent to apps.
Customer surveys are for companies who didn't pay proper attention to begin with.