HPE Ezmeral Data Fabric Review

​MapR-DB is a NoSQl datastore on top of MaprFS. The data can be updated and random data can be picked very fast.


What is most valuable?

MapR-DB is a NoSQl datastore on top of MaprFS. The data can be updated and random data can be picked very fast.  MapR-DB stores data in MapR-FS and it does not have region server like in HBase.

How has it helped my organization?

We have a couple of Hadoop development projects actively going on. Some of them are time series data and others involve customer interaction and marketing data. A NoSQL data storage is ideal for such applications.

What needs improvement?

It would be great if it were completely compatible with HBase.

For how long have I used the solution?

9 months.

What do I think about the stability of the solution?

No.

What do I think about the scalability of the solution?

No.

How are customer service and technical support?

Our customer purchased a paid support service and so far MapR has addressed our issues well. They even provided a fix for one of the ecosystem tools upfront, before putting it in the official release.

Which solution did I use previously and why did I switch?

Yes. We were using HBase for our previous solutions and one of our products. Our customer had already purchased the license of MapR, so the decision was not from our side.

How was the initial setup?

The installation and configuration was a simple process and it came with the product. They provided all required configuration setup scripts.

What's my experience with pricing, setup cost, and licensing?

The MaprDB comes with an "extract" cost. So if the requirements include Hbase, I would suggest going with HBase, instead of  MapR-DB.

Which other solutions did I evaluate?

Yes we prefered HBase, as it is distribution independent. All the industry major Hadoop distributions also support HBase. However, our customer had already made the decision to use MapR-DB.

What other advice do I have?

The biggest drawback is that it has vendor locking. It's proprietary in MapR distribution and extract cost. The positive side is that its storage is inherently within the MapR-FS and there is no region server overhead. Consider these aspects before making your decision.

**Disclosure: My company has a business relationship with this vendor other than being a customer:
More HPE Ezmeral Data Fabric reviews from users
Find out what your peers are saying about Hewlett Packard Enterprise, Cloudera, IBM and others in Hadoop. Updated: September 2021.
534,299 professionals have used our research since 2012.
Add a Comment
ITCS user
Guest
1 Comment

author avatarMSegel
Real User

There are issues with MapR-DB and HBase when it comes to full table scans and working with SQL languages.

The interesting thing is that in 6.0, they implement secondary indexing. This should improve performance when working with SQL languages.