Orchestrator Limitations

Has anyone run up against a max limit of bots against an orchestrator? We have a number deployed and we are thinking of releasing to our user community, but before we do, we want to make sure we’ve scaled properly. I don’t know if anyone has any real world examples of mass deployment and true sizing so we can cost appropriately.


Hi @dtetlow
Buddy usually the bot wont be released in mass deployment to my knowledge because that would lead to large impact when any issue occurs with bot in the production…fine any how that differs between the companies…
But usually it will be deployed one by one…and that is how agile works…in real time production environment
May be if its like bulk migration of robots from one Orchestrator to another orchestrator, proper SIT and UAT testing must done for each robot and should be ran in their respective machines to check their robustness and integrety
Then taking the cost in to consideration i would recommend you to consider uipath technical support or sales team for further details and the reason some may use nearly a set of orchestrator say three to four orchestrator based on their customer size…or some may have less than two or three…so its all based on the numbers and the type of robots been used… usually IT people would prefer more number of unattended bots and less number of attended robots, thus can include more development and production robots as they both include features of unattended robots mostly…
So kindly talk with these mentors…they can guide you on cost effectiveness…

Hope this would help you
Kindly revert for any queries
Cheers @dtetlow

@Palaniyappan I appreciate your detailed and quick response. I am not looking for the sales answer to my question. I am asking for real world examples of people who have deployed and exceed the limitations of a single Orchestrator and what their experience was. Yes types of processes play a role in determining scaling. Processes consuming more bot resources and availability are going to drive a higher utilization of both the bot and Orchestrator. But I am looking for real life experiences of implementers and generalizations about where they saw the cut off happening.


I can’t speak from experience as we’re just getting started in our travels… [10 Unattended in Production and 10 Unattended, 12 Developer in Nonproduction.]

Has some information of hardware requirements to speak to the number of CPUs/RAM you need for up to 250 bots per single Orchestrator instance. It also highlights adjusting the pool size to your SQL Server if necessary.

How many Robots are we talking? If there’s concern wouldn’t it be worthwhile to think about being scalability from the get go, throw an LB in front of orchestrator along with the Redis cluster and add Orchestrator nodes as needed?

If you’re an Enterprise customer… Should be able to reach out to your CSM to get feedback on your inquiry, they’ll either find out the information for you or potentially invite one of their architects to speak about it. You could also ask if their are any other customers that they’re aware of that has gone through this and see if they are willing to share information indirectly or directly with you.

I spoke to someone in support once and they said that in a test they run, a single Orchestrator could handle 10000 bots. Obviously you have to take hardware into account