HPE 3PAR StoreServ is a flash storage option, from Hewlett-Packard Enterprise.
N/A
HPE Nimble Storage
Score 8.2 out of 10
N/A
Nimble Storage was acquired by HPE in 2017. The enterprise flash array product line now goes by the name HPE Nimble Storage.
N/A
Nutanix Cloud Infrastructure
Score 8.9 out of 10
N/A
Nutanix in San Jose, California offers their software-defined Enterprise Cloud as a hyper-converged infrastructure solution. The Nutanix Cloud Infrastructure solution combines the Nutanix Acropolis virtualization solution, Nutanix AHV hypervisor (though Acropolis works with other hypervisors), Prism cluster manager, Nutanix Calm and Nutanix Flow server management, and is available on the Nutanix NX series of server hardware appliances, as well as third-party OEM appliances.
We evaluated the Nimble all-flash array as a competitor of the HPE 3PAR. During the evaluation, not only did the unit under-perform to the 3PAR, but Nimble was also acquired by HPE. We feel Nimble is a great product, but the 3PAR 8200 is a better fit in relation to our …
Management decided to have multiple hypervisor platform and Nutanix was the best in the market so far. The cluster management was so easy to manage, so no regret.
I've been a big Nimble fan and would consider it if I was deploying in a green field site (it always comes down to price of course). Nutanix are an existing supplier so for now it makes the choice easier. We did look at VxRail but again for the VDI solution Nutanix won out.
We made benchmarks and evaluated other storage solutions. But the HCI component and the integration from storage, computer, virtualization are what mostly made the choice clear for us.
3PAR is what we used to use and with Nutanix AOS it's added more feathers, like being web based and can be accessed anywhere using your phone when you are walking around.
HPE 3PAR StoreServ is a great storage device for both file and block-level storage needs. You can use it in place of SAN devices required by database servers. It provides an amazing compaction ratio and, hence, reduces overall storage footprint and cost. This is, however, slightly costly if your environment doesn't really see high transactions most of the time.
HPE Nimble Hybrid storage is very cost effective that can provide the same Latency, I/O and Bandwidth compared with any other All FLASH storage which is 3 times expensive than this. From a feature comparison standpoint of view, there isn't much difference between Hybrid Vs ALL Flash. HPE Nimble Storage uses ALL Flash disk at the front end to process the data thus by reducing the cost by using a regular SAS disk at the back end. Its the software and the algorithm that HPE Nimble StorageE uses to achieve this. having said that, the overall data reduction, de-duplication and compression is above the mark as what the sales team promised initially. Overall, the storage is performing well without any challenges.
For an organization that requires top-notch performance HCI, Nutanix is the best. You may start with 3 nodes and expand the cluster as required. The management through Nutanix Prism Central and Element was so easy that even a Junior Engineer was able to handle it. The Nutanix platform is not suitable for organizations with a small budget and fewer requirements for high-performance infrastructure, as the Nutanix solution itself is suited for enterprises.
We are using HP 3PAR storeServ 8200 for both our file level and block level storage.
I really love the amazing compaction ratio provided by HP 3PAR. Some of our disks have compaction ratio above 25:1, which is super amazing. You can definitely look for around 75% reduction of storage footprints.
It has a great user friendly web console "SSMC"(StoreServ Management Console). This is an additional bonus and can be used by even non-storage people like DBA and Network to take a look at storage performance.
HP Nimble is basically an all in one SAN with SSD cache or in some cases fully SSD based however the thing it does the best is its front end GUI and user management capability making it easy for anyone to ramp up on managing it quickly.
The Nimble SAN is also very easily managed when it comes to hardware management. Outside of some major internal parts such as the mainboard, almost everything in the system is duplicated so single failures never bring the system down. Power supplies and drives are easily swapped out.
Nimble has incredibly capable and easily accessible support that is available 24/7.
The price point on Nimble, though higher than piecing together a home made SAN out of spare Windows Servers, is perfect even for some SMB's.
One-click upgrades; whether it's hypervisor, firmware, disk or other updates. This feature has drastically decreased complexity and administration time.
Data Locality. Not all hyperconverged technology is created equal. When I first purchased Nutanix they were the only vendor (and as far as I know, still are) that made sure the storage a VM used was on the same host that VM was running on. Given a normal operating state, the [storage] network is literally only used for replication data.
They got rid of traditional RAID. Nutanix uses software to determine where a VM's storage should be written and replicated to. This dramatically decreases I/O when changing the number of nodes in a cluster, be it on purpose or during a failure scenario. Ex. adding a new node: If one uses RAID arrays then enough space has to be set aside to create a new array that includes the new node, then all the information has to be copied over, and the old array destroyed. RAID arrays do not grow and shrink gracefully so Nutanix has designed a better solution.
The Nutanix management interface was built on HTML5. No more flash headaches!
The system requires a physical or virtual system to manage upgrades and patches and for call home features. Most other systems have this integrated into the array. Just another piece of infrastructure to manage.
There are some system jobs that come pre-scheduled in the initial setup, but there is at least 1 job that should be added to this by default. When you delete storage, you also need to run a job to compact the CPG (a group where the luns all reside) to reclaim the previously used storage as free. This should be done by default at least monthly in the background.
There are 2 different GUI front ends that can still be used. The older version (3 PAR management console) is being phased out (has been for 3 years).
I'd like the GUI to include more information for some of the features such as replication data totals each night. You have to go to the command line for this.
It would be nice to have a feature built into the GUI that would show you the command line equivalent to get the same results you are seeing in the GUI.
Although the intial setup was easy, they could always improve on that portion. During my setup, I did have to do a lot of back and forth with research on their site as to what each setting was that I was setting up. They could have provided some sort of description for each field within the setup that would have made it easier to know what they were having us set up.
The one downside I have working with Nutanix is the sales team. They seem to try to add in extra goodies to sales quotes or push for extras that you don't really need and you have to tell them to take them out. Don't be afraid to push back on them.
Need to analyze sizing with sales team to ensure right sizing.
We are very satisfied with 3PAR performance and especially the IOPS (Input Output Per Second) is pretty amazing. It is easy to configure and doesn't require much knowledge of storage for Tier 1 support team.
Nimble is doing its job well and any issues that do come up cause the Nimble support team to alert us before we would potentially see an impact to our production environment. I do wish we could expand into the unused space in the CS210 shelf which is limited by what I assume is a marketing/sales strategy, but we will likely add shelves moving forward.
AOS definitely make our dev/test virtual environment management much easier than before. And the consolidation the test/dev environment from Azure and Cisco UCS, we have less need to transfer large amount of data between different hardware platforms which was very big challenge. To expand the capacity is very easy to archive as well.
Almost perfect, some hoops to jump through after major upgrade, but overall simple and effective. Our storage administrator really likes the integration with vmware as it makes his life easier. Also it was no trouble integrating it with our active directory credentials. The only issue we had was getting the plugin in VMWare going initially.
It's not out of the box easy, but once you get the fundamentals the steep learning curve flattens out and the processes to get things done and how it works becomes very apparent. It's wrapping the slight change in workflow from prior VM management methods took time to unbox and apply the Nutanix Cloud Infrastructure way
The performance is nothing short of amazing. This is an HCI solution, and as any all-flash HCI solution is amazingly fast, Nutanix AOS fills local IO requests until its local IO is saturated before reaching out over the network. This lowers latency substantially compared to vSAN.
With their phone home technology, we usually have a replacement part on its way by the time we wake up in the morning, or if the event occurs overnight. We have had great support when we had questions or events that required a call.
Any time I have had to contact support, they have always been quick to respond, and very efficient in resolving any issues. When an action has been required on our side for a fix, they have been very helpful in explaining step by step what was required, and when replacement parts have been needed, we've had them within 24 hours.
Our implementation team were great and worked with us and got the program up and running very easily. Every time we called post implementation we immediately talked to an Engineer, which is so unusual in dealing with companies. Everything they have promised they have full filled. I think their support is top notch.
IPv6 is needed for link local discovery. We do not have IPv6 configured on our network so the easiest way to get our nodes configured and discovered by foundation was to configure the IPv4 addressing within the node prior to trying to discover with foundation.
We still use NetApp storage enclosures at each plant for our routine file shares (CIFS) but I would love to swap them out for the HPE 3PARs as we use for our ERP app data. The 3PARS are simply faster and much better at near-instant replication. But even more importantly we can get the reporting that we need from the 3PAR within a few minutes, even custom reports. 3PARs are much easier to manage than NetApps in general and I think that the only place that they compete is in compaction.
We used a Dell EqualLogic PS Series storage array for many years and it worked well. A big issue was that upgrading firmware required system downtime and during one upgrade I was unable to bring the systems back up afterwards. I spent several days on the phone over a holiday weekend with Dell and VMware support in order to get my systems back up and running. It made me very nervous about doing firmware upgrades after that.
Nutanix integrates very well with Rubrik for backup and protection of the environment. Nutanix gave us simplicity and scalability compared to VMware and allowed us to extend our infrastructure into the cloud using EC2. One unified management pane for all our workloads, unlike VMWare.
We find that return on investment is probably a better metric in most cases.
ROI analysis is more than an exercise. Companies must outline what their future looks like, even if it’s vastly different from what they’re used to and comfortable with.
As good as your financial analysis might be, displacing status quo infrastructure has a lot of emotions tied to it.
