We have just built a new environment with 2 x V3240’s. Installation appeared all OK but system admins are experiencing problems with performance from hosts.
The environment is 100% Solaris 10 (SPARC) and connected via a 1Gb dedicated storage VLAN. Storage is presented over NFS to the Solaris hosts.
Copying a large file from local storage on UNIX server A to the NetApp volume provides good performance.
Copying a large file from the NetApp volume to UNIX server B local storage provides good performance
Making a duplicate of the file in the filesystem provides poor performance. This is 3 times slower than 1 and 2 above. E.g. volume mounted on unix box and standard cp command used to create a duplicate of the file.
We are struggling to identify the cause. The configuration looks OK and we can prove performance is OK on bulk read/writes described above. The end storage and SAN fabric show no issues.
Has anyone seen a similar situation on V-series in an NFS only environment?