Workbench Resource Requirements
Hi Everyone,
2 questions about Workbench:
- Is running a job in workbench (specifically one pulling from an FTP connection) more CPU or Memory intensive?
- Can you run multiple instances of workbench all under the same user credentials at the same time?
We are planning on running workbench on EC2 instances in auto scaling groups based on how many jobs we have to run. There are EC2 instances optimized for compute power as well as those optimized for RAM - we are trying to determine which would best suit running workbench jobs.
This will be easy to test once we scale fully but we'd like to architect for scale as opposed to using the age old "wait and see" method.
Thanks!
Best Answer
-
A Domo engineer was able to help me out with this one and the answer is:
I would say that it depends on the job. A job with many rows / columns is more memory intensive than a job with a small amount of rows / columns. The same is true with CPU usage. CPU is also impacted by the number of jobs running simultaneously. As far as memory resources are concerned, Windows will use disk space as memory if additional memory is needed (using disk space will slow down the job performance and increase the CPU utilization because disk I/O is slow compared to RAM I/O.) The reality is that modern computers are fast enough that the disk I/O time is negligible.
Hope this helps anyone who finds themselves asking the same question.
- Austin, Major DOMO @ 42 North Rx1
Answers
-
1
-
A Domo engineer was able to help me out with this one and the answer is:
I would say that it depends on the job. A job with many rows / columns is more memory intensive than a job with a small amount of rows / columns. The same is true with CPU usage. CPU is also impacted by the number of jobs running simultaneously. As far as memory resources are concerned, Windows will use disk space as memory if additional memory is needed (using disk space will slow down the job performance and increase the CPU utilization because disk I/O is slow compared to RAM I/O.) The reality is that modern computers are fast enough that the disk I/O time is negligible.
Hope this helps anyone who finds themselves asking the same question.
- Austin, Major DOMO @ 42 North Rx1
Categories
- 10.5K All Categories
- 5 Connect
- 915 Connectors
- 250 Workbench
- 459 Transform
- 1.7K Magic ETL
- 69 SQL DataFlows
- 476 Datasets
- 186 Visualize
- 250 Beast Mode
- 2.1K Charting
- 11 Variables
- 16 Automate
- 354 APIs & Domo Developer
- 88 Apps
- 3 Workflows
- 20 Predict
- 5 Jupyter Workspaces
- 15 R & Python Tiles
- 245 Distribute
- 62 Domo Everywhere
- 242 Scheduled Reports
- 20 Manage
- 41 Governance & Security
- 170 Product Ideas
- 1.2K Ideas Exchange
- 10 Community Forums
- 27 Getting Started
- 14 Community Member Introductions
- 55 Community News
- 4.5K Archive