Stop the networker daemons so that some of the indexes can be moved. In the SunOS 5 system, use /etc/init.d/networker stop. In the SunOS 4 system, use ps -ef | grep nsr and kill(1) the processes.
Find a file system with enough space to move one of the client's indexes. Only one of the client's indexes should be moved, not the networker server's index. To find the size of a client's index, go to /nsr/index/clientname/db and list the contents using ls -l. The database file can be large (possibly over 500 Mbytes).
Move the contents of a client's index to the other file system and check that /nsr has freed the space to use. You might need to unmount and remount /nsr, or even to reboot to designate the space freed by the move, as available.
After the space is available, restart the daemons.
Open nwadmin. Under Clients--Indexes, select a client and use Remove Oldest Cycle to free more space.
Use Reclaim Space to reclaim the space from the removed cycles. After a few of the old cycles have been removed, enough space should be in the file system to move the removed client's index back.
Stop the daemons, and move the client's index back to /nsr/index/clientname.
Restart the daemons. Remove the oldest cycles for the client that was just moved.
Tweaking of the browse policy and retention policy might be necessary to prevent this situation from happening in the future.
Otherwise, as long-term solutions, add more hard disk and run growfs, or move /nsr to a drive with more space on it.