Managing NFS and KRPC Kernel Configurations in HP-UX 11i v3 (762807-002, June 2014)

Modifying the Value
If the NFS client sends a very high number of commit calls to server, it can consume the bandwidth of
both the network and the NFS server. It can even reduce the NFS write performance. In such
environments, enabling this tunable can significantly reduce the commit calls over the wire. Due to
the reduction in the commit calls and having a dedicated set of kernel threads for sending those
commit calls, it might aid improving the overall write performance of the application.
2.1.47 nfs3_max_commit_threads
Description
The nfs3_max_commit_threads tunable controls the number of kernel threads that dedicatedly perform
commit calls for NFSv3 filesystems. This tunable is effective only for the NFS mounts created with the
nfs_enable_enhanced_commit tunable enabled.
Tested Values
Default: 8
Min: 2
Max: 32
Restrictions on Changing
The nfs3_max_commit_threads tunable is dynamic. System reboot is not required to activate changes
made to this tunable. However, the number of threads is set per filesystem at mount time only and
when the nfs_enable_enhanced_commit tunable is enabled. The system administrator must unmount
and re-mount each filesystem after changing this tunable. Only NFSv3 mount points are affected by
changing the value of this tunable.
Modifying the Value
Before modifying the value of this tunable, examine the number of files being written concurrently by
the application writes. If there are more numbers of files being written concurrently, increase the
value of this tunable. Also, the network bandwidth and the CPU and memory resources of both client
and server should also be sufficient to increase this tunable. This increase enables you to effectively
utilize the available network bandwidth as well as the client and server resources for sending commit
calls at a much faster rate. However, the total number of asynchronous threads for NFSv3 cannot
exceed 20% of the available nkthreads. NFS mounts fail if the mount command cannot guarantee the
ability to create the maximum number of threads for that mount point.
If the network has low available bandwidth, decrease the value of this tunable. This decrease will
ensure that the NFS client does not overload the network. Decreasing the value can impact NFS
performance because it limits the number of asynchronous threads that can be spawned, and thus
limits the number of simultaneous commit calls over the wire.
2.1.48 nfs4_sfh_boost_search
Description
The nfs4_sfh_boost_search tunable optimizes the search operation in the shared file handles list.
The NFSv4 mounted file systems maintains a list of all the file handles which are opened by
applications in a shared file handles list. Any file operation which involves creation, deletion and
updating of files involves searching of this list to get the reference shared file handle. The
nfs4_sfh_boost_search tunable optimizes the search of reference shared file handle in this shared file
handles list.