Capacity and performance planning

Scale your automation with capacity and performance planning. To plan your capacity and performance for Control Room Local, understand the requirements, limits, and defaults that determine the number of simultaneous bots, user sessions, and processing rates.

Overview

Review the information on the maximum number of concurrent bot deployments and executions for Local deployment.

Nota: For information about attended automations, see Capacity and performance planning at the end of this page. The rest of the information on this page is applicable only to unattended automations.

The entity count in the following list provides information only about the maximum number of entities that the Control Room can support irrespective of the machine resources. Deploying additional machine resources does not increase the entity count.

Control Room supports either of the following, for maximum number of concurrent bot deployments and executions:

1000 concurrent bot deployments and executions on Bot Runner sessions (typically in a production Control Room)
700 concurrent bot deployments and executions on Bot Creator sessions (typically in a development Control Room)


Entity Type	Count
Users	10,000*
Roles	5000*
Schedules	1500
Audit entries	10,000,000*
Lockers	100
Credentials	5000
Repository files (Including all file types in the Control Room Public workspace) Nota: Repository files include dependent files that you check in to the Public workspace (visible on the UI) and related files from the database (not visible on the UI). For example, `.png` files generated after an automation is recorded are stored in the database and are included in the repository files.	25,000*
Repository folders	9000*
Queues	30
Device pools	10
Elementos de trabajo	30 million

* Applicable from Automatización 360 v.21 and later.

Concurrent bot deployments and executions

Control Room supports up to 1000 concurrent bot deployments and executions across the Control Room cluster. These 1000 bots can be a combination of the same or different bots.

Nota: If you use different bots, because of the load, the compiler slows down the execution for the first-time deployment. For subsequent deployments, bot execution process might accelerate.


Control Room edition and configuration	Number of Run as users
One instance of Local	Maximum 1000
Three instances of Local with High Availability (HA) configuration	Maximum 1000

The bot deployment depends on the network quality and the size of data that needs to be transferred from Control Room to Agente de bot.

Therefore, in the above scenario when the network connectivity speed is increased for the three Control Room instances, the network throughput improves, so the completion time will become shorter but the maximum accepted concurrent bot deployments will remain the same.

Bot concurrent schedules

Scheduling considerations when running bots repeatedly.

When configuring bots to run repeatedly on a schedule it is important to make sure that the time between runs does not drop below the total time for deployment and execution of the bot. Otherwise, sequential executions of the bot might overlap leading to unexpected behavior.

For example if it takes two (2) minutes to run a bot, do not schedule the bot to run every one (1) minute. The previous run cannot complete before the next run begins.

With the reference specification, it is possible to successfully configure concurrent schedules.


Activity	Schedule data
Concurrent schedules	100
Run as user per schedule	10
Total concurrent bots	1000

Attended automations

Attended use-cases typically have bots that run for shorter periods of time (less than 10 minutes), are run frequently, and only run on one device at a time. Execution is also typically started from the Desktop Assistant.

These high frequency deployments require more resources on the Control Room and the database server.

Control Room resources required:


Environment	Maximum concurrent users
Single-node environment 16 CPUs 32-GB RAM	100
Cluster 3 nodes 16 CPUs 32-GB RAM	750
Cluster 5 nodes 16 CPUs 32-GB RAM	1500

Database server resources required: Minimum 8 CPUs and 16-GB RAM with an SSD hard drive, for example, AWS instance type m5d.2xlarge.