IBM Flashsystem 5000, a winning bet
December 24, 2021
IBM Flashsystem 5000, a winning bet
Score 10 out of 10
- IBM Flashsystem 5000
Overall Satisfaction with IBM FlashSystem
As system integrators, deploying [IBM] FlashSystem 5000 series on our customers is a cost/benefits decision in [the] first place. This product range of the series has a very accessible entry point (5015) and can scale up to the NVMe technology (5200) with fantastic performances still in a no-cost solution. We think these series are, with no doubt, the best in [a] class of the current market offering. We implement most of these systems at the production sites replacing older arrays while some of them are deployed at Disaster Recovery locations. Indeed, we configure some of these systems as backend storage of backup specialized virtual appliances in order to make them more flexibles in scalability terms. Also, the arrival of isolated snapshots this last year has become the most valuable feature on the 5x00 series of the range.
- Super simple administration across the full product line: no other vendor has the same interface for the FULL product line, which is just unbelievable from the storage admin point of view.
- The most flexible configurations capabilities of all the top storage arrays sellers. Usually, other vendors impose strong configuration restrictions sizing the arrays and combining different disk types, this is not the case with the FlashSystem products.
- Full line replication capabilities: believe it or not you can replicate data not only between [the] same range models but between any model within the full product line, from the modest FS5015 to the high-end storage virtualizer SAN Volume Controller.
- Currently, the protected snapshots are scheduled from the external software Copy Services Manager. That’s great when you have several machines but for customers, with just one storage array an internal scheduler will be a more practical option.
- There is a wide number of entry-level FS50xx (5010,5030,…) out there. These machines currently lack the protection capability and will be great to do whatever IBM can to give them at least a minimum possible protection feature.
- We think that the Data Reduction Pools have to improve [the] performance of entry-level arrays. Currently, we prefer standard pools in these machines that lack the hardware compression of her bigger sisters.
- Along these years, our customers have valued the zero business disruption and no worries as the most appreciated return for their business.
- The recently added ransomware resistance features are also increasingly appreciated because of the same reason: minimum disruption+fast recovery in case of attack.
- Also, the fact of having the same management software since the Storwize series time has saved a large amount of valued time and effort to the technician staff.
The product line is perfect for virtualization workloads of any size including VDI (that refers also to the entry-level machines), disaster recovery, or multiple sites replication (because all of them replicate in any scenario: LAN, metro area or WAN). If you need a real 100% data availability you can achieve it [by] implementing the hyperswap topology or as backend storage of the SVC in Stretched Cluster configuration, both scenarios will give you impressive data availability, we can confirm it. [On] the other side, they are not well suited in 'filer' scenarios because all the product line is 'block only' and will require a file server on top. That’s the case of object store-oriented projects, in this scenario you have better options without [a] doubt (at a higher price range).