What should I do to improve the performance of HP Vertica Pulse?

To optimize HP Vertica Pulse performance, you should modify the "jvm" resource pool to match your hardware so that Pulse efficiently uses available resources. You may need to experiment to find the optimal settings for your hardware and your specific workloads. However, the following steps should help you to get started:
  1. First, estimate the number of concurrent calls to Pulse and other Java user-defined functions that will be executed on your system.
  2. Second, reserve 6GB of RAM for each JVM session. Use the 6GB rule of thumb regardless of the specs of your system. For example, if your server has 64 GB of physical RAM, and two Quad-core processors (8 cores), then you want to reserve 48 GB for JVM usage.
  3. Use the command below to configure the JVM resource pool; this command has a cluster-wide effect. By default, the value of ‘plannedconcurrency’ will default to the number of cores, but you can set it to the value in Step 1 above. If you need 8 sessions but you don't have enough memory (48G), you will either need to reduce number of concurrent sessions or amount of memory per session.
alter resource pool jvm maxmemorysize '48G' plannedconcurrency 8; With this setting, Pulse and other Java UDXs will use 48GB of the RAM available on the cluster, with 16GB available for other Vertica processes. As mentioned above, you may have to experiment to find the optimal settings for your situation.

Leave a Comment

BoldItalicStrikethroughOrdered listUnordered list
Align leftAlign centerAlign rightToggle HTML viewToggle full pageToggle lights
Drop image/file