Dell VPLEX Reviews

• For non-disruptive migrations
• For replicating non-EMC vendor storage arrays using RecoverPoint
• High availability using VPLEX Metro
• Used as additional read cache for storage arrays

We have see performance improvement in our databases. We can use older arrays for archiving.

I have used VPLEX for over five years.

We have not had any stability issues.

While scaling VPLEX from 2-engine to 4-engine, you need to be careful with the design connecting to VMAX. In our environment, we encountered an imbalance on VPLEX BE ports. But, we resolved that issue with the help of EMC.

Technical support is exceptional.

We did not previously use a different solution.

Initial setup was a little bit complex when connecting to VMAX.

We evaluated IBM SVC.

Initial architecture is important. It will help you to scale-out the solution easily.

Clustering (Metro or Geo). It helps organizations to instantly replicate their data to the backup site, while making this totally transparent to the end users.

It enables our customers to instantly distribute their block data across two different sites.

Maybe setup a one-to-many cluster, instead of just a 1-to-1. If EMC could come up with a way to establish a cluster from one VPLEX to many VPLEX(s), that would allow organizations to be able to have more than one copy of their data, and therefore have one, or many, backup (or disaster recovery) sites.

Currently, a cluster can only be established between 2 VPLEX nodes. For example, if a customer has 3 sites, he can’t replicate data from Site #1 to both Site #2 and Site #3. Therefore, with VPLEX, only a 1-to-1 replication is possible.

If there was a way to set up a one-to-many or many-to-one replication, then it would allow customers to have more than one backup site.

I have used VPLEX for three years.

We have not had any stability issues.

We have not had any scalability issues.

Technical support is excellent.

We did not have any previous solutions.

The procedure is very clear and easy to follow.

We evaluated EMC RecoverPoint.

It’s a great product; very reliable. Just get trained and understand all the concepts around EMC VPLEX.

The ability to move disks online without disruption to the environment.

Ability to create geographically distributed clusters.

Possibility to replicate volumes with storage devices from multiple vendors.

The production environment needs improvement.

I have used VPLEX for five years.

We experienced increases in latency in accesses to disks and some bugs on engines.

I did not have problem with scalability.

There were delays in solving problems.

I have two solutions: HPE 3PAR and VPLEX.

It is very laborious to allocate volumes on servers.

The VPLEX, along with the RecoverPoint price, can raise the price a lot.

Due to the large number of models and different brands it is a very good solution to virtualize volumes.

Evaluate the native features and a possibility of a good integration with the RecoverPoint.

Volume replication in almost real time, across metro distances.

With the help of EMC VPLEX, we improved the architecture of the geographic redundant infrastructure our application required.

Cost. This is really expensive.

I’ve used VPLEX for six months, during the architecture and implementation phase of the project.

I was in contact with presales support only, which did a great job.

We did use a previous solution, but it had a totally different architecture. With the help of EMC VPLEX we simplified the architecture and improved the response time.

The setup is pretty complex, since the solution is stretched across two sites at metro distances.

At the time of the project, I was not aware of any similar product.

Cross datacenter automated failover.

My organization uses two data centers. In the past, we would have to manually failover between those data centers. This timing would be unpredictable since it required human interaction. With VPLEX, we are able to automate this process where failovers between sites now occur automatically.

The licensing model for enterprise storage (EMC VMAX) and integrating the solution into the actual storage frames instead of a separate product. I do believe Dell EMC is addressing both of these concerns in current and future product offerings.

I have used EMC VPLEX for three years.

We have not had stability issues.

We have not had scalability issues.

The solution just runs and is very stable. It is only after an issue, such as a site disruption, that we even contact support to validate everything is healthy.

We didn’t have a previous solution.

The initial setup was straightforward. There was a bit of a learning curve on our part since this was our first implementation of a storage virtualization product.

Our current licensing model is mid-tier arrays (EMC VNX) that are licensed based on frame, which lends itself to putting all servers attached to that frame under VPLEX control. The enterprise arrays (EMC VMAX) are capacity based. This makes using VPLEX on these frames very expensive and we only do the systems that are most critical and require failover between sites on these frames. Dell EMC has told me they are addressing this concern.

We did not evaluate any alternatives.

We used it to primarily provide automated failover to our VMware virtualized environment. I would highly recommend any organization get their virtualization team involved right from the beginning. Depending on what you are trying to accomplish, you may have to change how you setup your VMware environment to accommodate the VPLEX implementation.

Based on our requirements, we bought this product for doing online data migrations without data disruption, but we realized that there is a limit to backend storage that can be encapsulate to it. Hence, we ended up spending more money of buying additional VPLEXs to support more backend storage, which in turn came out as an expensive solution for us.

We don't see any improvement in the organizational functions.

If it can support more backend storage per VPLEX.

I have used EMC VPLEX for three years.

We have had stability issues. It adds another layer of connectivity and hence another layer of single point of failure.

We never went into a situation of upgrading/scaling out the engines on the VPLEX.

Our main motive was to perform data migrations across the data centers from one environment to another and that is why we bought this product. But we ended up doing database and host-based data migrations due to the complexity of the product, which adds another layer of connectivity. For a new application setup, it worked like a charm.

It was straightforward to deploy the VPLEX, but the trick was the encapsulation of existing live applications to the VPLEX. It would required downtime.

It’s a bit of an expensive solution if you have multiple storage arrays that need to be encapsulated.

We were recommended by EMC to use this product for our dynamic environment where we do a lot of data migrations in and out.

To move LUNs non-disruptively between arrays and there is a mirrored LUN between data centers, that is to read/write on both sides which are very helpful features.

The business continuity is sorted, as the data is in two data centers, i.e., readable and writable. If one side goes done, there is no downtime and the other side continues non-disruptively.

It needs to improve the thin reclaim on certain arrays.

I have used this solution for three years.

I have not encountered any stability issues.

I have not encountered any scalability issues.

The technical support is very good.

Initially, I was using IBM SVC. It was a bit different and it was not a true active-active solution.

The setup was a bit complex in order to split the traffic and the zoning.

There are two license types, i.e., Metro and Local.

Make sure this is what you need.

The review concerns EMC VPLEX product, which can be purchased/used in two different ways :

1) VPLEX Local

2) VPLEX Metro.

VPLEX Local is used to virtualize the storage locally, within the same data center.

With VPLEX Local, one possible use case is to mirror data between two local arrays. In this way if we lose one array then the data is still accessible to production server on the second local array.

With VPLEX Metro the data can be mirrored between two distant arrays, each one installed on a different data center. With this configuration even if we lose an overall data center (disaster condition), then the data is still available on the other site. This is an additional level of data security which permits to face up disaster conditions.

Another use case offered by VPLEX (either Local or Metro configurations) is transparent data migration.

Without VPLEX, the storage migration (or data migration) can be done in two ways:

- Host based : the host manages the data migration, consuming from its own resources (CPU, RAM etc). A downtime can be needed as part of the host based migration.

- Array based: migration is handled by the array, but a downtime has to be scheduled host side to discover the new storage.

With VPLEX we can migrate data from one array to another, or within the same array while a server is accessing data.

I’ll give four examples:

1. VPLEX Local configuration: We did experience a serious issue where a storage array needed to be rebooted on a production environment. Fortunately this array was encapsulated under VPLEX, and so we were able to migrate data to another array transparently to users, then reboot the problematic array. Finally we migrated back the data from the “spare array” to the source one. During all these operations no user was aware of migrations, and no downtime was consumed.
2. VPLEX Local configuration: A new array was purchased then encapsulated under VPLEX. The new array had more performance than existing ones, so a data migration to the new storage was recommended from application owners to ameliorate performance. The data migration was done transparently and without application downtime.
3. VPLEX Metro configuration: From a disaster recovery point of view, VPLEX Metro offers more security to data. We didn’t encounter a disaster yet, but we did test the failover of applications to a second site during implementation and the test was really convenient. I would also recommend adding RecoverPoint to the VPLEX Metro solution to add more security to production data.
4. VPLEX deployed for RecoverPoint: We have EMC RecoverPoint that we use to replicate data from non EMC arrays to a VNX. In this case we have to encapsulate the non EMC array under VPLEX then use VPLEX Splitter to replicate data. Without VPLEX, use of RecoverPoint on non EMC arrays would not be possible.

Much room for improvement exists. For example:

• Storage tiering cannot be handled by VPLEX, as it is performed by back end arrays. It would be great if we can configure tiering between arrays!
• When configuring a mirror (RAID1) device on two devices A & B, performance of the mirror depends on A & B individual performances. The response time of this mirror will be equal to the lowest response time of the A or B devices. It would be preferable that the mirror response time is equal to greater performance of single components, by setting a preferred access path for example (this feature exists on IBM SVC).
• No native data service exists on VPLEX (I mean snapshotting, replication features, etc). We have to buy other products such as ViPR to be able to use those features.

We have used this solution for four years.

We have not encountered any stability issues.

We have not encountered any scalability issues.

The level of technical support is 10/10. Over four years we didn’t encounter any issues with VPLEX technical support.

SVC was used locally. VPLEX Metro has unique capabilities so we switched to VPLEX Metro as we had to implement a two active data center architecture.

The initial setup was not really complex. The only difficulty was to plan downtimes for each platform, as the initial encapsulation requires a new storage discovery from the host perspective.

Regarding licensing, VPLEX uses capacity licensing. I think before purchasing the VPLEX you must study your data growth over the next five years and from the beginning order the overall capacity licenses. Extensions as much more expansive than the first purchase.

We determined that no other products offered VPLEX Metro capabilities in 2012. The only possibility to implement Metro features without VPLEX is to use Symantec Volume Manager software, but this is a software solution, without cache options and with more licenses to purchase, and more deployments to implement. VPLEX Metro was the only solution for our targeted architecture.

VPLEX offers a large compatibility matrix and valuable features. The technical support is also appreciated for this product so I would encourage others to purchase it.

VPLEX would not be your best solution, only if you want to implement tiering between arrays. If you use VPLEX the storage tiering will be handled on back end arrays. Other use cases can be implemented with VPLEX, individually or with other EMC DELL products.

Also if you plan to use RecoverPoint Replication features on non EMC arrays, then you have to encapsulate those arrays under VPLEX. So if you have a plan to deploy a Disasater Recovery Plan within a few years, you should start to think about it now. If RecoverPoint is your preferred solution and you have non EMC arrays, purchase VPLEX from the beginning.