Provides reliable performance but could benefit from better deduplication management and intuitive admin toolsRegarding the NVMe over Fabrics feature of Hitachi VSP 5000 Series, we tested it, but there was no admin tool available for configuration until about a year ago. This was problematic for us, so we decided not to use it until we have the proper management tools. The main difficulty with this feature is that it's not supported for GAD devices, so Global-Active Devices are not supported with NVMe. That's basically what all our ESX platform is based on. We believe we would get the most benefit if we could use it in the ESX platform, but it's not supported yet. The deduplication impacts the quality of block storage that system administrators were accustomed to, resulting in increased service time due to the high CPU usage. The whole deduplication should be handled better in the next Plattform, should not have such a high impact on contoller CPU usage. The admin tools, called Ops Center Suite, have been inadequate for years, which Hitachi acknowledges. We had to find other vendors to provide a simple, fast GUI to provision storage. There are limitations, such as a storage pool only having 1024 pool volumes, which means the storage pool is maximum three petabyte in space. This requires building multiple storage pools, resulting in multiple pools where you have to keep some space as reserve. Having more than one storage pool forces you to have multiple times some percentage of storage capacity as a reserve, that you paid for but can't use it.
