hi all i have got one issue saying that /ora filesystem 98% full...so.pls tell the steps to know the reasons for this and solution

Typically it would mean that 98% of the unreserved space is utilized. Since you didn't tell us what reported it or the actual message it isn't clear. For example some utilities also report on inodes and space (e.g. Nagios plugins). So in a filesystem with thousands of little files you might have free "space" but be running out of inodes as the each file uses a separate inode (except for hard links within a filesystem).

First check the filesystem to be sure space is the issue:
df -h /ora

If space isn't the issue try checking for inodes:
df -i /ora

If space IS the issue then you should first look for memory dumps called "core" files:
find /ora -name core -type f
If you find any core files type "file core" on whatever directory they're in to verify they are core dump files. Typically these will take up a lot of space and can be deleted unless you're planning on doing analysis of the dump.
***WARNING*** Oracle creates DIRECTORIES as opposed to regular files that are also named "core". Be sure NOT to delete any oracle core directory.

Another thing that tapes up a lot of space is tar files. Doing a gzip of those will give you space back without having to delete the file.
find /ora -name "*.tar"

You can drill down on the directory:
du -sk /ora/* |sort -n
That will show you what directories/files are taking up the most space with the largest being at the bottom.
If it is files you can check the individual file to determine if it should be there and be that large. Often you'll see logs or log directories as a leading culprit. If the logs are not currently in use you can often either gzip them or delete them.
***WARNING*** Deleting a log or other file that is currently held "open" by a process does NOT free space. It deletes the name only and the space is still in use.

Another common situation is that presented by the last warning above. Someone has deleted a file that was held "open".
To determine if this is the case you can do an lsof /ora and see if there are any CHR files that have no name associated. The only way to clear such files is to stop/kill the process that has them open. The PID will show up in the lsof output.
***WARNING*** Please be sure you know what the process does before you attempt to stop/kill it. It may be doing other very important things that make doing planned downtime a better idea.

Of course rebooting a server will get rid of all processes so if you suspect the last scenario is your problem and can't figure out which process has a file open a reboot will solve it.