QPR ProcessAnalyzer System Requirements: Difference between revisions

From QPR ProcessAnalyzer Wiki
Jump to navigation Jump to search
Line 49: Line 49:
Dataset sizes don't affect the QPR ProcessAnalyzer Server requirements, because data is processed in Snowflake.
Dataset sizes don't affect the QPR ProcessAnalyzer Server requirements, because data is processed in Snowflake.


== Recommended Snowflake Configuration ==
== Recommended Snowflake Warehouses ==
The more there is data, the more calculation nodes are needed for the Snowflake virtual warehouses. The more there are simultaneous users, the more clusters are needed for the Snowflake virtual warehouses.
The virtual warehouse sizing in Snowflake follows these general guidelines:
* Scaling up: The more there is data in the process mining model, the larger warehouse size is needed (XS, S, M, L, XL).
* Scaling out: The more there are simultaneous users or the more there are charts and analyses in the dashboard, the more clusters are needed for in the multi-cluster warehouse.


Typical node size configuration:
Typical Snowflake warehouses for different model sizes:
* Models up to 10M events: 1 node (XS)
* Models up to 10M events: 1 node (XS)
* Models up to 30M events: 2 nodes (S)
* Models up to 30M events: 2 nodes (S)
Line 59: Line 61:
* Models up to 1B events: 16 nodes (XL)
* Models up to 1B events: 16 nodes (XL)


Typical clustering configurations:
Typical multi-cluster configurations:
* Up to 2 simultaneous users: 1 cluster
* Up to 2 simultaneous users: 1 cluster
* Up to 4 simultaneous users: 2 clusters
* Up to 4 simultaneous users: 2 clusters
Line 66: Line 68:
* Up to 20 simultaneous users: 5 clusters
* Up to 20 simultaneous users: 5 clusters


It's also possible to setup Snowflake to set the number of clusters dynamically based on the load.
It's possible to set Snowflake to adjust automatically the number of clusters based on the load, which is the recommended configuration. There are two scaling policies available and of them the standard scaling policy is suitable for QPR ProcessAnalyzer analytics usage.
 
* More information about warehouses: https://docs.snowflake.com/en/user-guide/warehouses-overview.html
* More information about multi-cluster warehouses: https://docs.snowflake.com/en/user-guide/warehouses-multicluster.html.


== Recommended Hardware for QPR ProcessAnalyzer Server (in-memory) ==
== Recommended Hardware for QPR ProcessAnalyzer Server (in-memory) ==

Revision as of 00:08, 10 February 2023

System Requirements for Web Browsers

One of the following web browsers are needed:

  • Windows desktop
    • Google Chrome
    • Mozilla Firefox
    • Microsoft Edge
  • Mac desktop with Safari
  • Tablets and smartphones
    • Apple iPad and iPhone with iOS 16 and Safari
    • Android version 9 or higher with Chrome for Android

Note: Due to the rapid update schedule of web browsers, QPR Software does not specify the supported versions but indicates the latest versions that are tested and verified to work. Chrome and Firefox versions newer than the tested versions are likely to work, but QPR Software cannot guarantee their functionality before the next release. Any compatibility issues will be fixed as soon as possible for the next release.

System requirements for QPR ProcessAnalyzer Server

System Requirements for QPR ProcessAnalyzer ScriptLauncher

System Requirements for QPR TaskRecorder

  • Operating systems: Windows 10 or Windows 11
  • Installation uses approximately 100MB of disk space, and additional of 200MB should be reserved for task recordings.
  • Approximately 200MB of memory is used by QPR TaskRecorder.
  • QPR TaskRecorder is a .NET 6 application released as a self-contained package, and .NET 6 doesn't need to be installed separately.

Recommended Hardware for QPR ProcessAnalyzer Server (Snowflake)

When running QPR ProcessAnalyzer in Snowflake, there are following requirements to QPR ProcessAnalyzer Server:

  • Up to 5 simultaneous users
    • 16GB memory
    • 4 processor cores
  • More than 5 simultaneous users
    • 32GB memory
    • 8 processor cores

Dataset sizes don't affect the QPR ProcessAnalyzer Server requirements, because data is processed in Snowflake.

Recommended Snowflake Warehouses

The virtual warehouse sizing in Snowflake follows these general guidelines:

  • Scaling up: The more there is data in the process mining model, the larger warehouse size is needed (XS, S, M, L, XL).
  • Scaling out: The more there are simultaneous users or the more there are charts and analyses in the dashboard, the more clusters are needed for in the multi-cluster warehouse.

Typical Snowflake warehouses for different model sizes:

  • Models up to 10M events: 1 node (XS)
  • Models up to 30M events: 2 nodes (S)
  • Models up to 100M events: 4 nodes (M)
  • Models up to 300M events: 8 nodes (L)
  • Models up to 1B events: 16 nodes (XL)

Typical multi-cluster configurations:

  • Up to 2 simultaneous users: 1 cluster
  • Up to 4 simultaneous users: 2 clusters
  • Up to 7 simultaneous users: 3 clusters
  • Up to 10 simultaneous users: 4 clusters
  • Up to 20 simultaneous users: 5 clusters

It's possible to set Snowflake to adjust automatically the number of clusters based on the load, which is the recommended configuration. There are two scaling policies available and of them the standard scaling policy is suitable for QPR ProcessAnalyzer analytics usage.

Recommended Hardware for QPR ProcessAnalyzer Server (in-memory)

  • Small datasets (up to 5 million events per model) with simultaneous users
    • 32GB memory
    • 4 processor cores
    • 50GB disk
  • Medium datasets (10-20 million events per model) with simultaneous users
    • 64GB memory
    • 8 processor cores
    • 100GB disk
  • Large datasets (20-80 million events per model) with simultaneous users
    • 256GB memory
    • 16 processor cores
    • 150GB disk
  • Very large datasets (more than 80 million events per model) with simultaneous users
    • 512GB memory
    • 32 processor cores
    • 200GB disk

Recommended Hardware for QPR ProcessAnalyzer Database Server

There are two databases used by QPR ProcessAnalyzer: metadata database and datatables database. The metadata database has relative small amount of data, so pretty lightweight SQL Server is sufficient (for small installations, 4GB memory and 2 processor cores, and heavy usage environments, 8GB memory and 4 processor cores). The datatables database has similar requirements, but in addition the required disk space depends on amount of data stored (100GB - 10TB SSD disk).