Data360 DQ+

We have new hardware for Hadoop cluster and DQ+.  I'm going to be installing DQ+ in 4 new clusters (will be decommissioning existing virtual environments).  We have SND, DEV, USR & PRD. 

In these environments, SND and USR will have 4 nodes to install and configure DQ + components.  So the proposed configuration will be the following:

Node1: Application, Standby-ApplicationDb

Node2: Application, ApplicationDb

Node3: Application, Load-Balancer

Node4: Vertica

In the DEV and PRD environments, we will have 5 nodes to install and configure DQ + components.  So the proposed configuration for these two environments is as follows:

Node1: Application, Standby-ApplicationDb

Node2: Application, ApplicationDb

Node3: Application

Node4: Load-Balancer

Node5: Vertica

Are there any issues or concerns with these proposed configurations?  

Hi Josh,

I reviewed this with our Infrastructure team, and the number of nodes looks good. In the outline below, we shuffled some of the components between the nodes. The reasoning was to isolate components consuming more resources or to pair large components alongside low resource consuming components.

4-node setup:

Node1: Application Server 1, Load Balancer, Vertica Node 3

Node2: Application Server 2, ApplicationDb, Vertica Node 2

Node3: Standby-ApplicationDb, Vertica Node 1

Node4: Vertica Master

5-node setup:

Node1: Application Server 1, Vertica Node 4

Node2: Application Server 2, ApplicationDb, Vertica Node 3

Node3: Vertica Node 2, Load Balancer

Node4: Vertica Node 1, Standby-ApplicationDb

Node5: Vertica Master