The following is a guide on the minimum recommended setup needed to run Pyramid. This does not mean it is the recommended setup.
These recommendations are slightly different for Kubernetes deployments.
|
Item |
Setting |
Comment |
|---|---|---|
Server |
|
|
|
Host Operating System |
Windows:
|
64-bit only. Desktops should only used for basic testing. |
|
Linux:
|
64-bit only. Access to repo's are required. |
|
|
Host CPU’s |
Minimum cores: 8. Recommended Minimum: 16 |
Production environments should provide ample CPU's to process queries. See suggestions below. |
|
Host Memory |
Minimum memory: 8 GB Recommended Minimum: 16 GB |
Production environments should provide ample memory to host the application, in-memory databases as well as the needs of the operating system itself. See suggestions below |
|
Host Disk |
Minimum space of 2 -3 GB |
Use of very fast disks like SSD's are highly recommended |
Client |
|
|
|
Client Browser |
Any fully compliant HTML5 browser (e.g. Chrome, Edge, Opera, FireFox, Safari, IE11) |
The client browser can operate on any operating system, as long as it can connect to the web application hosted on the relevant server. IE11 works but is not recommended. |
|
Client Screen Resolution |
|
The desktop client must support the minimum screen size to operate effectively. |
Hardware & Environment
The right setup is heavily affected by the specifics of each deployment. There are countless variables that need to be quantified and considered. The guide to scaling Pyramid provides an in-depth review of these items. Briefly, some of these considerations include:
- Number of concurrent users : This variable will heavily influence the number of CPU’s needed to keep performance optimal.
- Query styles and consumption: If your users typically run very large queries (5,000+ rows) and do this consistently, the underlying hardware should provide enough CPU cores to not only process such large responses, but also provide enough bandwidth for other users while such queries are being processed. Networking speed will also have an impact as more data is streamed to the end users.
- Size of the databases and the types of technologies they operate in: In-memory databases will require lots of memory to host the relevant databases. On-disk technologies are less sensitive to memory, but will require VERY fast disk access. OLAP usually needs both, but does not require the same memory footprint or amount of disk space. Generally, In-memory and memory based OLAP engines require strong CPU's.
- Up time and back-ups: Multi-Server deployments are best suited to ensuring the highest up-time, which in means that both the hardware strategy for the application as well as the location and performance of the server hosting the database repository is important (which remains central).