What is most valuable?
Storage virtualisation: Use o storage pool to spread I-O load across as many disks as possible, reduces “hot spots”.
Ability to manage third-party arrays and virtualise them: One screen to control multiple arrays. Simplified administration.
Auto-tiering with Easy-Tier: High usage volumes and parts of volumes are automatically moved to high performance disks in manageable “chunks”. This means volumes with heavy I-O usage are are moved to SSD for the best response. Zero admin overhead.
How has it helped my organization?
Using V7000, the storage administrator has complete visibility into the I-O load and response times of the volumes, storage pool. Together with “Easy Tier”, this means volumes with high workloads are automatically moved to the highest performing disks (SSD) in the array. There is almost zero administration required once the storage pool is established.
Data migration: The migration wizard made migration from old to new an absolute doddle. Something that historically had been “hard and difficult” became a background task, once the old array was managed by the V7000.
What needs improvement?
The replication of data between IBM Storwize arrays in V1 of the product had some major limitations. A lot of the limitations were overcome in V2 of the product. Some improvement can still be made in scalability using clustered V7000 nodes.
One area for improvement is in the GUI, where host clusters are not properly dealt with. With Hypervisor host clusters, all hosts must see the same volumes in the same order. Using the concept of a “host-group” has been around (even with IBM) for many years, so why not with the V7000?
For how long have I used the solution?
What do I think about the stability of the solution?
Some isues were encountered during upgrades in the early firmware versions. This has now been resolved.
What do I think about the scalability of the solution?
The V7000 can scale up to 10 enclosures (disk shelves), each up to 24 disks per enclosure. A cluster of 2 x V7000 has twice the capacity.
Using the V7000 you must be careful not to overload the CPU and disk backplane. When planning to implement any storage array, you must plan your workload carefully. IBM have a superb tool, Disk Magic, that predicts the I-O performance of the array configuration you have selected. This is an invaluable tool in planning any IBM storage array.
How is customer service and technical support?
I was lucky to have a very good relationship with the local IBM storage support experts. The implementations were done under IBM guidance and, knowing this, any issues that came up at my sites were quickly dealt with. The support staff knew the arrays had been installed with expert guidance and had no inherent issues in design.
Which solutions did we use previously?
Previous storage solutions included the HPE 3PAR array family. We switched to IBM for 3 reasons:
- Upgrade cost
- Complexity with administration tasks and day-to-day running
- Support costs
How was the initial setup?
The IBM V7000 initial setup was straightforward, as I had engaged IBM support early on and was given very good advice on how to size, configure and implement the array(s), including data migration from the old to the new.
What's my experience with pricing, setup cost, and licensing?
It pays to go back and get the best price you can from your supplier. The first offer is not always at the best discount.
Ensure that all the features you need (including replication) are included in the license options.
Which other solutions did I evaluate?
Several vendors were evaluated in addition to IBM: HPE, NetApp, EMC.
What other advice do I have?
- Download the IBM Redbook on best practices to implement.
- Engage IBM support or your integration partner/reseller at a detailed level.
- Know what you want to do and document your requirements.
- Investigate upgrade costs and five-year maintenance options.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Sep 17 2017