Cloud Management Platform vRealize Operations

How to Gain Full Monitoring into Hadoop Clusters

By: Lora Johnson

Because of its flexibility, scalability, resiliency, and power, Apache Hadoop is the big data solution of choice for many organizations of all sizes. However, one of the challenges that users face comes with visibility.

The size and organization of the distributed database makes it difficult — if not impossible — to get the insight necessary to understand how your Hadoop clusters are performing. And because of that, if issues do arise, it can be challenging at best to pinpoint where they are coming from, whether from other areas of the IT stack or from Hadoop itself.

Whether you are running Hadoop on VMware or bare metal, Blue Medora has a new solution to give you the visibility that you need to gain insight into your Hadoop clusters. With the Blue Medora Apache Hadoop Management Pack for vRealize Operations, clearly understand the capacity and health of a HDFS cluster, and ensure the memory and CPU devoted to clusters are optimally allocated. And if you are planning that virtual migration, you can use the management pack to quantify the real resource needs of your cluster.

Key Performance Metrics

With the Blue Medora Management Pack, users get insight into the health and capacity of HDFS NameNodes and DataNodes, as well as the health, capacity and performance for YARN ResourceManagers and NodeManagers.

Access over 400 metrics, including system CPU load, used physical memory, bytes and capacity used. In addition, the management pack incorporates key metrics for containers and applications so you garner true visibility into resources associated with Hadoop.

Capacity Definitions

Metrics are one piece of the Hadoop monitoring puzzle. Take it to the next level with capacity definitions — which provide health and utilization breakdowns so you clearly understand how clusters are being used and where you may be able to improve performance.

Go beyond seeing a green indication that all systems go, and dive deeper to ensure you don’t overload particular Hadoop clusters — reducing mean-time-to-innocence and improving troubleshooting across the IT stack.

Dashboards and Reports
With a series of out-of-the-box dashboards and reports within the Blue Medora Hadoop Management Pack, see key metrics in at-a-glance heatmaps and graphs to quickly evaluate performance issues.

In all, there are six dashboards available within the management pack, which is complemented by three reports, including the Hadoop HDFS Capacity and Health, and the Hadoop YARN Utilization and Health.

How to Get the Hadoop Management Pack

The Blue Medora Hadoop Management Pack is now available. To learn more about the management pack for vRealize Operations or download a free trial, please visit the True Visibility Suite for VMware vRealize Operations page on Blue Medora’s website.