All-Flash Arrays Forum

Content Specialist
IT Central Station
Jun 14 2018
One of the most popular comparisons on IT Central Station is Dell EMC Unity vs NetApp All Flash FAS. One of the users on our site says about Dell EMC Unity, "As long as you know what you're doing storage-wise, Unity is really easy to use." Another user says about NetApp All Flash FAS, "I think that the most important thing is the integration with the existing features that we already have, and existing management systems." In your opinion, which is better and why?
reviewer70797Both the Unity and NetApp all-flash arrays are very good all flash solutions and work well. I would not be able to recommend one over the other but would suggest the customer look a little deeper into what requirements they have for the storage such as cloud tiering, setup, integration with other products, management and make the determination based needs.
rrukhaiyar573Unity is still not that reliable .. as it has so many bugs in the software...plus UEMcli is not that user-friendly. Hybrid models are good on load bearing capacity but overall has still a long way to go. Have known Netapp Solidfire since the days it was only Solidfire... and the IOPS guarantee marketing regime is something next gen. I would rate Solidfire above Unity and NETAPP FAS arrays also above Unity models.
Alfred Morgan JonesCompanies looking for a an all flash system will be swayed by 3 main factors: What storage vendor they already have, price and how well the proposed solution is marketed/sold. Feature comparison normally comes about 5th on a list of the most heavily weighted factors. In my many years of advising clients it has been my experience that if the IT manager knows and likes a particular brand, that is the one they will purchase. So, a client with Netapp will buy Netapp. From a features viewpoint: Both have a lot of very useful features and both have horrible "gotchas" that can trip up unwary administrators. If you buy a new brand, spend money on training !!
Content Specialist
IT Central Station
Jun 14 2018
One of the most popular comparisons on IT Central Station is EMC VNX or HPE 3PAR Flash Storage. One user says about HPE 3PAR, "The optimization features move chunklets or hot spots to faster drives." Another user says about EMC VNX, "The replication feature provides another functionality to protect your data." In your experience, which is better and why? Thanks! --Rhea
Chaan BeardActually the answer is neither. VNX and 3Par are both technology best described as yesterday's hero. Look instead to another HPe product called Nimble A series AFA or the AccelStor AFA series. AccelStor are actually delivering frightening performance in their P710-SLED platform which delivers over 400K Mixed load IOPS, has 69TB effective capacity and is fully VMware VAAI integrated and certified for an MSRP of $49,995.00. Either are so easy to install 8 year old kids can do the job without much head scratching going on. AccelStor does not charge by capacity and comes with all the software you need baked into the price and this includes snaps, clones, mirrors and full replication capability. HPe 3PAR and Dell EMC VNX platforms come with capacity based licenses and replication pushes the costs up through the stratosphere. Between Nimble and AccelStor for the reporting capability the Nimble is very nice but you also pay extra for those nice reporting capabilities, but it is good. AccelStor gives you what you need to know, no nonsense style - basic reporting but they are working hard currently on new features that bring them up to par with Nimble's fantastic reporting engine. 3Par and VNX are not true All Flash Array platforms by the way, they are Hybrids that accommodate both spinning disk and All Flash SSD. Nimble A series and AccelStor's entire range are purpose built all flash arrays with operating systems designed to cater for all flash only. www.accelstor.us is their website. Enjoy!
reviewer325866As far as I know, VNX line is no longer available, therefore I think that the real comparison has to be with the Dell EMC Unity family systems
James MercerBoth are fairly old technologies, and frankly speaking, replication is nothing new at all. As another individual stated here, if you want to examine something up-to-date that completely eclipses both 3Par and EMC, HPE Nimble will demonstrate a clearly superior approach to flash and non-flash storage.
User at a tech vendor with 1,001-5,000 employees
Mar 08 2018
How does HPE 3PAR Flash Storage differ from INFINIDAT InfiniBox? Which do you recommend?
Content Specialist
IT Central Station
Jan 23 2018
Does anyone have experience with both NetApp AFF and Pure Storage? A number of users recently inquired about how they stack up, what are the reasons you might choose one or the other or the differentiating aspects of each product.  How do NetApp and Pure compare?
Jean Carlos Bitencourt Da SilvaHello, In the design of choice of our AllFlash storage we tested several equipments, among them NETAPP and PURE. The finalists were: PURE / NETAPP / EMC. All delivered the performance and features that we had as success requirements. But ... our choice was PURE for simplicity of implementation and management, in addition to its innovative licensing model where you have the right to use everything that is implemented and what will be implemented yet you will also have right of use. We have been with PURE for 12 months and very satisfied with the results.
Bunyamin TasdemirHello, We don't use AFF but we used 8020 SAS and 8020 SSD and we replaced them to PURE. IO problems minimized, volume limitations resolved I degraded servers ram and cpu which were working on Netapp but when i moved systems to PURE, they started working as never degraded
reviewer503028I don’t have a NetApp AFF array. Our NetApp is actually quite old so I’m sure that taints my experience but from that experience, I will say it’s night and day difference between Pure and NetApp in ease of use, support, and customer service. Pure is set it and forget. It’s been mostly hands-off since we purchased it which for our team was a big selling point. I believe the NetApp AFF we were looking at still had some quirks with how patching was done even down to updating firmware on the individual drives. The NetApp allowed for more tuning of the pools whereas the Pure managed it all behind the scenes. In a larger environment, those may be selling features for the NetApp but those things were detractors for us as we don’t have a dedicated storage admin. Hope that helps.
Senior Manager with 1-10 employees
Jan 22 2018
Hello Everybody, Can someone help me understand the differences between oracle ZFS5-2 and Oracle FS1-2? Thank you,
Sr. Director of Community
IT Central Station
If have an RFP/RFI template or evaluation matrix for choosing an all-flash array, please share it with the community in order to help others out. If you email it to me (ariel@itcentralstation.com) we will have it cleansed and remove all company and private information before posting.
Manager Communications and Infrastructure at a financial services firm with 1,001-5,000 employees
Which do you recommend? Dell EMC XtremIO Flash Storage or Hitachi Virtual Storage F Series?
Jeff DrennanI would personally take a look at Nimble Storage. It is all processor and memory based. You can get the same performance out of a Hybrid array from Nimble. They offer all flash as well, but performance is not based on disk I/O. Very cool technology.
Infrastructure Manager with 1,001-5,000 employees
The IBM FlashSystem V9000 compression is a lot higher than EMC XtremIO and Hitachi.  IBM claims to be around 4.5:1, whilst EMC and Hitachi guarantee around 2:1 for Oracle workloads on AIX.  The IBM guarantee is for 90 days from implementation whilst the EMC guarantee is there for the duration of the maintenance. How true is this IBM compression figure for general Oracle on AIX?
reviewer127386To IT Central Station and prospective all-flash (AFA) buyers: Thank you for reaching out on this topic, these vendor’s claims are confusing to many non-initiated buyers. On face value, it appears that some technologies could perform much better based on their messaging. I posted a blog on this topic here: https://community.hds.com/community/products-and-solutions/storage-systems/blog/2016/08/01/many-all-flash-array-data-reduction-guarantees-are-sounding-like-us-republican-presidential-nomination-politics Note that the basis for vendor’s data reduction claim do vary greatly as some vendors choose to include benefits from thin provisioning and snapshots in their factoids (aka: alternative facts). Keep in mind that any “up to” reduction value message is just that; a value achieved in lab or unique workload and not representative of average. In this case, both EMC and Hitachi choose to represent average compression results from prior deployment; no deduplication value is included. IBM provided best case as this V9000 model does not support deduplication; consequently, it creates the impression that this V9000 system delivers superior result with deduplication supported! The variation in data reduction (compression and/or deduplication) result is mainly a function of the data set, not the vendor’s technology as engineers are limited to similar latency overhead. Here is a sample of typical Hitachi Storage Virtualization Operating System (SVOS) compression and deduplication results by workload that Gartner validated: [cid:image002.png@01D27726.4E691DA0] Hitachi benchmarked internally both IBM and EMC AFAs on eight different workloads and data reduction results came in within 5% of each other’s. Data reduction services performance will also be a function of the data chunk size; as an example, a 4KB chunk size engine can achieve about 5-10% better results versus an 8KB engine. Note that this extra savings will come at the performance and memory cost. So how do you proceed as a prospective buyer to assess the value of each technology? I recommend using vendor’s data reduction estimator tool on your own data set and make your own conclusion. Hitachi estimator tool can be downloaded here: https://hcpanywhere.hds.com/u/9xzMSp4czIPlHMxK/hidr_estimator-02_1_0.zip?l ; data reduction estimators are built-in with the similar algorithm as their AFA counterpart. As for performance impact, I would recommend reading this blog: https://community.hds.com/community/products-and-solutions/storage-systems/blog/2016/11/03/flashbuyer . In the real world, you can’t have your cake and eat too… Full disclosure, I do work for Hitachi Data systems and support Hitachi enterprise flash business. I hope this information helps. Patrick Allaire

Sign Up with Email