IBM FlashSystem Reviews

I provide IBM FlashSystem to my customers and host it on the GBM private cloud, to give them the best services. I am working on a project to replace an HP Primera with an IBM FlashSystem 5200. This solution consists of two boxes connected with the remote mirroring feature of the FlashSystem. Each box has about 30 terabytes of storage and is replicated using a flash copy. The boxes are connected to a Power System running AIX, and I have implemented the EC tier with multiple layers of disk including MV, SST, and rotational disk. Additionally, I have set up Spectrum Protect to make backups to the TS 4300 tape. This project is currently being implemented and should be completed within the next month.

At the FlashSystem level, customers are especially fond of multi-tier and distributed rate systems, particularly the dynamic rate six arrays. They appreciate the extra layer of protection this provides, as well as the Easy Tier feature which is analogous to IBM’s FlashCopy. Customers find the FlashCopy feature highly beneficial, and the feature often comes up in conversations.

Customers are delighted with the new version of IBM's Virtualize system because it allows them to easily expand arrays with a single disk, something that was not possible in the past. They appreciate the added convenience of being able to incorporate multiple items indistinctly into the array.

Our customers have raised concerns about the limitations of the FlashSystem 5200 and 7300, which only offer a 32-gigabyte connection. This presents an issue for customers whose other infrastructure, such as old HBA cards, is limited to an eight-gigabyte connection. To maintain the recommended N minus one connection, the customer must use the 32-gigabyte connection, which is not ideal. To provide more flexibility, our customers would prefer to have a 16-gigabyte connection on these small boxes. The customer was initially unhappy with the situation because they had to replace all of the eight-gigabyte connections in their sixteen servers in order to migrate to a new storage system. Furthermore, to remain competitive, the customer had to be given the machines for free. While the customer accepted this, it was still a displeasing and uncomfortable experience for them. However, the machines were eventually sold and installed, leading to a satisfactory outcome. Despite this, the customer was still left with a feeling of dissatisfaction from the negotiation.

I have been using the solution for several years.

I give the stability a nine out of ten. The latest versions of Virtualize, from 8.5 and above, are highly reliable. I have not had any complaints from customers about them.

I give the scalability a nine out of ten. This solution is highly scalable, allowing for both side-by-side placement and clustering. This makes it an ideal option for me due to its scalability.

My technician told me that the initial setup of this machine is much faster and easier than with previous versions, and they were very pleased with it. Though I haven't experienced it myself, I understand the positive experience it has provided.

For example, with the small model, we can configure the machine to enable us to access the SAN. This includes configuring the worldwide name and getting familiar with the SAN experience. It typically takes around three to four hours to set up boxes in the morning, depending on the size of the array and any specific configurations that the customer wants. For smaller models, this is usually the case. For larger models, the setup time may vary but is estimated to take no more than four hours for arrays up to 30 terabytes. 

For deployment on the small machine, one to two people are typically required, as it is necessary to have two people to rack the machine. For larger machines, three people are usually used for racking and setup. The large machines are the DS, the Enterprise, and the 7300.

I'd rate the basic licensing and the Virtualize software a ten out of ten, and the extra Spectrum and other an eight out of ten. However, if we combine the basic and advanced licensing, I'd give a score of eight out of ten.

I give the solution a nine out of ten because of the solution's flexibility, scalability, and great prices. Last year's performance was remarkable, and I expect the same this year. IBM FlashSystem truly exceeded my expectations.

I suggest using the ESS for planning and ensuring a thorough conversation with the customer about their desired outcome. Proper planning is essential for successful implementation; if the planning isn't done well, it can lead to having to redo the work. My advice is to plan carefully and talk to the customer to make sure we understand exactly what they need and how to deliver it.

The solution has good performance and the highest throughput ratio in a single enclosure. The most valuable feature of the solution is SCM (Storage Class Memory), which has the lowest latency value in the storage industry.

The solution's infrastructure technology level could be PCI Express 5 instead of PCI Express 4 for the next version.

In addition to SaaS-based enclosures, the solution's expansion options must include NVMe options.

I have been using the solution for a few months.

I rate the solution a nine out of ten for stability.

The transactions of more than 1,00,000 customers are running on this solution.

I rate the solution an eight out of ten for scalability.

We have previously worked with several store systems like Pure Storage, HPE, and NetApp storage. We switched to IBM FlashSystem 9100 NVMe because of its performance.

I rate the solution ten out of ten for the ease of its initial setup.

The solution’s deployment took around two days.

The solution has good pricing, but it is not cheaper.

Overall, I rate the solution a nine out of ten.

My company hosts the database from Oracle on IBM FlashSystem.

The most valuable feature of the solution is replication since, in our company, we also use SVC, which is a different product from IBM. In some sites, we can replicate the data from SVCs back to FlashSystems, which is the most valuable feature of the tool, considering the technology that we use.

In IBM FlashSystem, data reduction is an area with shortcomings where improvements can be made in the future.

In the future releases of IBM FlashSystem, I want to see synchronization in three-site replication specifically.

The response time taken by the technical support team of IBM is an area where improvements are required. It takes a lot of time for me to get to IBM's second line of support to get an answer.

I have been using IBM FlashSystem since 2016. I use the solution's latest version. I am a customer of IBM.

Stability-wise, I rate the solution a nine out of ten.

I never try to scale up the solution. I mostly buy a lot of small boxes of the solution to operate in my company.

In my company, two administrators at the most use IBM FlashSystem.

I have a lot of interactions with IBM's technical support. I rate the technical support an eight out of ten.

Positive

The solution is deployed on an on-premises model.

The business partner provides my company with the service related to the solution's initial setup.

As per the policy of my organization, we usually use three alternatives at the same time. In our company, we compared IBM with Pure Storage and Dell. Each product is used for a different workload in the same area.

I rate the overall tool a nine out of ten.

The Flash core models offer amazing performance.

The basic setup can be challenging when it comes to certain IP addresses and the configuration of the IP. You have to go in to different menus to makes changes and ensure it is stable.

Ina future release, we would like to have a better way to install and to configure FlashSystem.

I have been using this solution for two years. 

I would rate this solution a nine out of ten for stability. I have experienced one problem and that was solved by EBM. It was for a FS 5,200. I had to connect to an IBM i system without some switches. I was the second or third customer who did it, worldwide. They had to make a special patch for it.

This is a scalable solution. 

I would rate the technical support for this solution a three out of five. 

Neutral

I would rate the pricing of this solution a four out of five. 

We considered Dell PowerStore. From a capability and technology point of view, both solutions are comparable. However, from a price perspective, IBM is 30% cheaper.

I would definitely recommend this solution to other users. It is suitable for many sizes of business due to the options they offer. 

I would rate this solution a nine out of ten as it offers great performance and stability. 

Our company uses the solution for central storage. Every person in our company is served by and accesses the solution. 

We also resell the solution to small business customers.  

The solution is very easy to configure and use. 

The solution's pricing is a bit high. 

I have been using the solution for eight years. 

The solution is very stable so stability is rated a ten out of ten. 

The solution is scalable so scalability is rated an eight out of ten. 

Technical support is great so is rated a ten out of ten. 

Positive

We sell the solution to customers who handle their own setups. 

The price is a little bit high so is rated a six out of ten. There is always room for improvement on the solution's price. 

I recommend the solution and rate it a ten out of ten. 

We implement this solution for our customers, including African banks who use it for storage. 

The storage capacity of this solution could be improved. 

I have been using this solution for six years. 

This is a scalable solution. 

The technical support for this solution is good. They used to help us when the motherboard of Power Systems broke. Their response times are really fast. 

The initial setup is straightforward but does have some complexity in the middle. 

The price depends of the technology that our customers need.The price can come at a lower cost but this may increase as storage is added.

I would recommend this solution because it offers good performance and the systems are fast. In terms of dealing with Oracle, it's really a good choice for customers that have their infrastructure that needs balance. 

I would rate this solution ten out of ten. 

Our primary use case for this solution is managing and supporting our workloads currently running on IBM FlashSystem. Additionally, we use it to run our ERPs.

We like how straightforward the IBM system is and the technical team's support.

IBM FlashSystems is lagging in optimizing storage technologies, which can be improved. For example, deduplication with Dell EMC and their storages are far ahead of the curve regarding data storage deduplication. Deduplication is when you remove all the data duplicates from your system and save more data in a smaller physical space. If there are any duplications in the entire storage, they should be removed, and data can be saved. There are two types of deduplication, inline and post-save deduplication. Inline refers to the data travelling from the system to the storage to be stored. The system removes the duplicated part and then stores it to save space. Another kind is that you store the data first and then remove the duplications, so you will need a larger area to read and write. The inline deduplication needs to have global deduplication enabled.

We have been using this solution for five years.

The solution is stable and has no issues.

The solution is scalable and has varying degrees of scalability. For example, if you have to buy a port and initial or entry-level storage, it can scale up to a certain level. After that, you'll have to change the box. If you need a forklift to pick up and bring in a new device regarding scalability, we usually plan for three to four years because technology changes fast. Hence, with five years of scalability, you will have no issues scaling that product to a certain level. 

I rate customer service and support ten out of ten.

Positive

IBM storages are straightforward. The only issue is the higher-end machines of the V9000 storage systems. There are two ways storage can be configured, and if you have multiple controllers in storage, you don't need to enable zoning between the two controllers. However, for IBM V9000 and above models, the controllers are also zoned through your zoning system. I rate the initial setup an eight out of ten.

We have no issues with the price as it is very competitive. So, if you want to buy high-end IBM storage, HP would go higher than IBM, so if it's a million-dollar storage and you have around 150 terabytes of hybrid or all-flash storage that you deploy, then it costs approximately 1.7 million dollars. This solution is the second most expensive in the market.

I rate the solution nine out of ten. The solution is good, but variable-length deduplication and global deduplication can be included.

We used the solution exclusively for block storage. Over time, it added compression features and now even NVMe. 

It's perfectly suited for an on-premise solution or for providing a base for cloud solutions, VMware workloads, IBM i-series, IBM AIX, IBM Power, Linux, and Windows compute. In other words, the complete server stack. It is something others actually can't offer. All of this can be operated from within the same solution. 

It definitely has a strong plus in environments where you actually have such different server solutions in place.

It's a really reliable, powerful platform. It's a mature product. It's like a BMW that evolves consistently. 

There is no need to change or buy another company's solution. It came with storage virtualization and options to move/migrate volumes around and migrates easily even before you actually have svMotion on VMware. 

It can be stretched. There is a Site Recovery Adapter. It has backup integration using flash copies. You can build a disaster recovery solution around it. IBM has its famous Redbooks where you can enter in the best practices. You name it, they've got it!

It offers separate IO Modules for connectivity additions, for compression to offload the CPU. 

It offers storage virtualization to ease migrations. You can build storage clusters and migrate data and easily configure partner relationships. 

The solution offers excellent performance! Flashcopies come in handy with backup solution integrations. The site recovery adapter for VMware Site recovery manager integration is great. Everything is working like a charm. 

I've used it in a banking environment in combination with VMware Site Recovery Manager and Site Recovery Adapter (SRA) - a wonderful combination. It saves you headaches building a recovery plan. Most of all, it works.

IBM's solution has come a long way and has had different milestones/features have been introduced. I would position the 7200 in the upper midrange class as it has lots of features - more than, for example, EMC Unity/VNXe. It doesn't lack anything, really. One could argue that NAS or S3 is not available to it, but I prefer other types of storage optimised for that job.

It is slightly more expensive, however, it all depends on your supplier. Licenses are volume-based. Larger companies with more TB usually are better off as the price per TB decreases the bigger you go. I would definitely recommend this platform!

I've known the predecessors of the IBM Flashsystem, being the Storwize 7200 (Gen 1 to 3) and SVC Front End Servers, from the time they were running on code 6.2 way back in 2011.

We have never ever had an incident with it making the infra go on its knees, nor did we have any datacorruption ever.

All storage solutions have bugs and all have their correctives that might not address an issue on the first occurance or under all circumstances. When the system is stressed and under some specific (nonetheless rare) conditions , the developped code might trigger a reboot of a controller node  to avoid data corruption. A system with 2 controllers is resilient enough on its own and a reboot of a node to prevent e.g. cache merge problems or alike is not harmfull.

To me, the need to evict a controller node and warm boot it is actually intentionally a safety precaution that avoids data corruption, something we all want to stay out as much as possible. It's as reliable as any other product in that respect. All solutions that I know off like DELL EMC Unity, Fujitsue DX-series or 3PAR Storeserv respond in the same manner to avoid datacorruption. I've seen it the most on the SVC (code 7.1/7.2 around 2012/2013  , but not on the  Storewize v7000 Gen1 to Gen2+ solutions, though I have to admit they all had about 40 to 50% of the load of the SVC.

You can build clusters with it and IO Groups.

IBM is able to deliver a support organization with well-trained people. IBM's redpieces/papers and solutions designs are published and offer real references. You can develop your own skills and become an expert or fallback on IBM Supreme Support if you feel less comfortable. 

Positive

We used IBM FastT/DS Storage (LSI Logic) with/without IBM SVC and IBM v7000 Gen1/Gen2/Gen3, shifting from legacy/traditional storage to storage virtualization.

It's quite intuitive. I always enjoyed that, whatever operation you initiated on its Webvinterface, it generated the command, and you actually see which command is being executed. I love it! It surely helps you get acquainted with the product. The setup with many modules and ports requires some design, as on any midrange solution. You get a nice template to initiate it and off you go.

I've done it myself after all the experience I've had. 

At the end of the lifecycle, the migrations and options you have mean that it doesn't require several storage solutions if you have mixed server solutions like Mainframe, i-Series, AIX, IBM PowerPC, VMware, Desktop Virtualisation, et cetera.

Also, the superb integration with VMware SRM and its Site Recovery Adapter makes it a seamless solution to make your infra resilient.

It's not as expensive as HPE Storeserv, however, it is more expensive than Netapp FAS or Dell Compellent. 

We looked at the Netapp FAS all-flash series, Dell Compellent, HPE StoreServ (7200/7400, 8200/8400, and 10500), and Dell EMC VNXe/Unity

It's a more complete solution and really up to mixed infrastructure and resiliency and has a lot to offer on scalability too