NFS Session Delays
The NFS session delay exceeds the critical value set in Expert Thresholds. The "session delay" metric combines network latency (i.e., the time spent waiting for data to traverse the network) and application processing time (i.e., the time spent waiting for the server or application to respond to a request). Application Response Time for NFS is tracked with a separate Expert Threshold.
If the Expert Threshold for application processing time is also being exceeded, the slowdown is probably the result of an NFS client/server configuration issue. Perhaps the server hardware is overloaded with disk I/0 requests, or some other system component (NIC, processor) is too slow given the traffic it is expected to handle. There are entire books available that describe performance tuning the various versions of NFS, but some of the things you can immediately check are:
The size of the cache buffer, both on the client and server side.
The maximum block size allowed by the server, and the block size specified by the client. This configuration option can also affect the network latency portion of the delay.
The performance of the underlying (i.e., non-NFS) file systems. For example the content and layout of the server's local file system can greatly affect NFS performance. Directories with large numbers of files, or large numbers of symbolic links to NFS auto mounted file systems compromise the server's ability to respond quickly to requests for data.
If application processing time is within reasonable limits, network latency must account for the bulk of the session delay. You should look for problems with the network. Do you have enough bandwidth between the application server and the clients that use it? Are the routers between the systems configured correctly?