In present environment, broadly, there are five types of database that are being used more frequently. They are traditional, NoSQL, In-memory, Hadoop and columnar databases. There are others as well like Virtualized or federated Database, streaming databases etc. From EMC’s perspective, the good news is that we have solutions to the 5 top variants as mentioned before. In this blog, I am going to discuss about the NoSQL databases which act ,sometimes, as a complimentary entity for traditional databases like Oracle. NoSQL Databases can manage the exponential increase in the volume of data, objects, and products and also the frequencies at which these components are accessed coupled with high performance, scalability and accessibility. NoSQL databases like MongoDB can handle the following issues with greater efficiency compared to the traditional databases.
Under present circumstances, we are very fortunate to learn that EMC has a solution around Mongo DB and XtremIO storage which is ranked No. 1 in All Flash Array segment. Initially, Mongo DB was running on DAS(Direct Attached Storage), but to avoid risks and costs to the organizations and to gain better performance and scalability , MongoDB has proved to be much more beneficial when used with XtremIO. By using this solution we get the following benefits:-
Let us now do a deep dive into the solution (of MongoDB with XtremIO) with the help of Figure 1, 2 & 3.
Figure 1 : Hardware Configuration of MongoDB with XtremIO storage
Figure 2 : Server Configuration
Figure 3 ; Software Configuration
In the solution, the above components help in accelerating the scalability factor and performance that can be taken to multiple nodes with the linear increase in IOPS bandwidth and capacity coupled with sub-millisecond latency as is demonstrated in Figure 4 below.
Figure 4: MongoDB and XtremIO IOPS and Latency
The above figure demonstrates that under no circumstances the performance of the MongoDB databases will degrade as the above solution is maintaining the sub-milli-second latency throughout the life of the databases. The solution has also achieved an overall capacity efficiency of 22:1(as demonstrated in Figure 5 below). Also we see in Figure 5 that out of 27.083 TB of used capacity, XtremIO has only consumed 3.98TB of physical capacity. We can also see that 7x storage capacity requirements got reduced with XtremIO features like inline deduplication & compression. It is needless to say that more we save on capacity and storage , the more we save on our company's IT expenditure and budget.
Figure 5: Data Reduction Ratio and other Efficiency parameters of XtremIO with MongoDB.
MongoDB with XtremIO helps not only the DBAs and storage administrators in terms of storage reduction and performance but also enhances the productivity and turnover of developers by utilizing XtremIO’s snapshot services, in-build advance agile processes and full-performance snapshots. Also, different MongoDB environments like test/dev/SIT and production can be consolidated under one umbrella --> XtremIO Storage. XtremIO’s snapshot services help to refresh production environments with parallel environments, thereby generating speed for copies and maintaining SLAs. Lastly, MongoDB workload snapshot consolidation with the aid of XtremIO improves TCO and accelerate MongoDB workflows. The snapshot operation is basically done via XtremIO GUI as shown in Figure 6.
Figure 6 : GUI of snapshot copy of Mongo DB in XtremIO Storage
I have to stop here and leave out some more good features of this solution in the interest of brevity of the blog. In summary, we can derive the following benefits from MongoDB on the top of XtremIO Infrastructure:-