Pivotal Knowledge Base


How to Choose GPText Instances Per Host in GPText 2.x


Pivotal GpText 2.x


This article describes how to choose GPText instances per host in GPText 2.x.


In GPText 1.x, it will require one GPText instance for each gpdb primary instance. But in GPText 2.x, as the product design was different, we don't require one GPText instance per each gpdb primary instance. Generally configure one or two GPText instances for one host is enough. But there are some exceptions.

Here is a quote from one GPText dev engineer:

GPText could run multiple instances on one machine, but for a typical installation, this is not a recommended setup. Extra CPU and memory resources are required for each additional instance.
A single instance or a small number of instances are easily capable of handling multiple indexes.
For every recommendation, there are exceptions. For the recommendation above, that exception is mostly applicable when discussing extreme scalability. The best reason for running multiple Solr nodes on one host is decreasing the need for extremely large heaps.
When the Java heap gets very large, it can result in extremely long garbage collection pauses, even with the GC tuning that the startup script provides by default.
Because of the potential garbage collection issues and the particular issues that happen at 32GB, if a single instance would require a 64GB heap, performance is likely to improve greatly if the machine is set up with two nodes each having a 31GB heap.
So for example, if the instances have the 16GB memory for GPText, one node with 16GB JVM is enough. If the instances have more than 32G memory forGPText, it's better to set up more than 2 nodes.
The GPText1 uses solr4. Many new updates have come since we use solr6 in GPText2.



Powered by Zendesk