IBM Elastic Storage Server vs. IBM Storage FlashSystem vs. Scality ARTESCA

Reliable data storage and access is a big challenge in AI applications and IBM ESS is a service you need to solve the problem. If you are building an AI service and you already use the IBM ecosystem for AI compute requirements, ESS can serve as an excellent software based storage solution. Integration with other platforms such as EMR/Databricks is still a challenge and should be improved.

Incentivized

Top choice for IBMi integration. In general, it's a very good solution for quite a lot of customers. If you need a 3 site replication solution (HA + DR) maybe it's not the best solution, unless you use SVC. FlashCore Modules have very powerful compression rate, while mantaining good performance: high IOPS rate and really low latency

Incentivized

The team in charge of implementing Scality has to be full stack in order to guarantee the correct functioning of the entire system. However, you have to think very carefully about the balance between servers and disks, perhaps adopting smaller fully populated servers instead of large semi-populated servers, which would mean that over time our disk updates will not have a fully useful life.

Incentivized

• Powerful processing power, literally support unlimited scale out storage capabilities
• well integrated and supported with multi flavor environment and file systems.
• Huge capacity with in a very reduced form factor of disks/enclosures

• Extremely fast storage
• Scales up as needed. Add flash core modules easily and increase volumes very quickly.
• Applying firmware updates to IBM Storage FlashSystem is very easy to do and can be done without any downtime.
• Replication policies are very versatile and also easy to setup.

Incentivized

• I love the redundancy of the software.

Incentivized

• Integration and configuration need a bit more technical expertise which is I guess expected.
• There can be some improvement in integration with Legacy applications.

Incentivized

• Definitely we need to be able to extend a volume that is part of a mirror pair or Policy-Based Replication(PBO) on the fly.
• Expansion of mirrored volumes/volume groups should resynchronise only the changed data (Almost like making use of change-volumes) and not trigger a full resync.
• Please allow us to select downloading drive microcode also from Fix Central directly onto the flashsystem like with updates.

Incentivized

• The overall packaging is not very good. I think it could be more efficient for installation.
• There are many components in storage servers. This can generally be complex to understand, you have to be patient.

Incentivized

As flash technology evolves rapidly, vendors should [adapt] to the changes and onboard them in their products. IBM is a bit slow in adopting advanced technology in [its] portfolio. Even though the current product is good, my decision is purely based on how the transformation of the product [is] being produced and supplied to [the] end-user.

Incentivized

The integration into our SVC and the midrange systems are very well working. We do use cross site replication and global mirror which in turn is easy setup and the consistency groups are well maintained. The storage does global mirroring but also internal cross room replication as well. The integration of cloud object storage also helps in tiering the overall system.

Incentivized

As an organization, it took us a while to understand the shift from a traditional black box SAN to software-defined storage, but now we are much more certain of what this means. The time invested and the resources were not very high, thanks on the one hand to the technical support and on the other to the coherence and good development of the platform.

Incentivized

Yes. The last really issues I've seen in the environments of my customers were back in the "Storwize" era of systems. Issues with code updates or sometimes rarely failing critical components. BUT: since every component is at least dual redundant even back in that day real outages were very rare. And today in the "FlashSystem Era" the systems are so reliable and available even during code updates there are no complaints. I think the 5 or 6 9s are completely achieved.

Incentivized

Latency and throughput have been excellent. The compression is done natively on the NVME modules so we do not see an impact there. the distributed parity does not seem to impact write speeds at all. We have found database performace has increased quite a bit over our old hybrid storage solution.

Having a single tier of performance makes it easy to set up SLAs for workloads.

Incentivized

I would rate the overall support for IBM Storage FlashSystem with a 10. The response times have consistently been fast, and the support engineers always come prepared with relevant questions and practical solutions. They demonstrate a deep understanding of the product and our use case, which means issues are typically resolved quickly and without unnecessary back-and-forth. This level of support has been especially valuable for our SAP environment, where uptime and quick resolution are critical.

Incentivized

I have seen Scality in the office meeting with our VP and get the feeling that they are here to support us.

Incentivized

Very good traning by experienced personnel. I attend instructor led sessions on IBM TechExchange and STUs before as well as IBM workshops for FlashSystems at IBM Böblingen which I helped to prepare and improve after piloting them. Absolutely nothing to complain about FlashSystem training in Germany or USA. So: all good here!

Incentivized

THe training is often selfpaced and easy to follow if you have worked previously with other storage products. The content is clear and contains all the information needed for an admin to learn. However the troubleshooting part is not always included and this is dependent on forums or other sources where the information is not accessible very easily.

Incentivized

I was amazed by the way IBM Flash storage provides a huge amount of storage within the limited footprint, it helped us to replace big fat storage arrays, which aided in the process of moving towards lower hardware, carbon, and energy consumption footprint. We can comfortably say that Flash helped us to host applications fast with a blazing speed and user transactions are quite quick.

Incentivized

IBM ESS is optimized for AI and Big Data usecases while S3 is a general purpose storage solutions. EMR and Databricks have lakehouse/data warehousing solutions for distributed computing but are more optimized for just the big data pipelining solutions and not essentially for AI usecases, especially for inference, when you need to load model artifacts really quickly.

Incentivized

For the past six months, we have been utilizing this technology in our projects to manage a massive workload and produce swift business outcomes on a crucial production database. The IBM FlashSystem 7200, which offers high performance and compression, has so far been a great experience for us. The results and functionality of this product have delighted our consumers.

Incentivized

Due to the nature of our business we require extensive encryption and availability for sensitive customer data. The erasure encoding that Scality provides gives us the assurance that documents are rest are never in a state of being downloaded or available to a casual data thief. This is something that can be found with other vendors but at a fraction of the same cost. Having this kind of performance, availability and redundancy at the cost that Scality provides has made a large difference to our organization.

Incentivized

I'm using cognos - especially cognos reports - as a part of IBMs Storage Intelligence software to manage datacenters and report storage and overall resource consumption to my customers. The software is reliable and - my customers who are using cognos in their businesses - say that they are very satisfied about the scalability of cognos. I can give it only an 8 rating because of the quite heavy resource footprint...

Incentivized

• When the cost and price dynamics comes into account, we have replaced age old traditional arrays with almost negligible cost with ESS
• able to extend the capabilities of global teams to work and share their experiences and knowledge with other teams.

• the compression in line means the IBM flashcore allow us to store much more information with less capacity, we can store much more with fewer disks
• virtualization allows us to extend the useful life of old equipment that we already had separated, as I said, I use them as a repository for backups
• With the implementation of the new functionality of immutable copies with SafeGuarded Copies, the information is much more secure for possible attacks

Incentivized

• Keeping sensitive customer data secure is a must for our organization and Scality has great features to make this happen.
• We replaced a single SAN with a Scality ring and found performance to improve as we store more and more customer data.
• Being able to lose various portions of our Scality ring and allow it to continue to service customers while maintaining high performance has been key to our business.

Incentivized