How to Trigger a Bot from the Control Room API |  

We're happy you find these videos useful, Jim!

Hi Anish. At this time, there is no API to stop an in-progress bot.

Thanks for the feedback Theo! The docs page for the Control Room Deploy API is currently being updated - it has better explanations of mapping different data types now: docs.automationanywhere.com/bundle/enterprise-v2019/page/enterprise-cloud/topics/control-room/control-room-api/deploy.html

Thanks for watching, PS! Any application where API calls can be made could be used to trigger a bot in your environment. That could be through custom JavaScript, a custom Java application, Python script, or any number of other program languages/platforms that enable connections to other applications through API calls. If you look in Postman, the code that we used to trigger the bot can be exported in any number of languages directly from the Postman tool.

Hey Robin - no separate link for a Postman collection...though many of the end points are the same/similar to what was shared for Automation 360.

Hi Gabriel. We recommend that you check which role botrunner1 is a part of and make sure that your account has access to trigger bots via this runner.

Hi D. The second task for processing would be queued up to execute after the initial task is completed.

Hi K Van. Thanks for the enthusiasm! Authentication tokens default timeout is 20 minutes - though this could be configured to a longer time period. Alternatively, using the /authentication/token endpoint would enable you to refresh an already-valid token on demand - further extending your time. When you think of the ID’s to log in to the CR API and bot runner accounts - those are 2 separate things. An account that is set up only for the purposes of kicking off automations would not have to be (and likely shouldn’t be) the same account used for running the bot runner. Consider creating some accounts that only exist for the purposes of running different jobs via the Control Room API while separating those functions out from any Bot Runner accounts. Should you need additional clarity on how this could be done, consider posting your scenario to A People or opening a support request ticket.

@@Automationanywhere Thanks for that added info...That all makes sense and we'll consider just that for our ID's!

Hey States - the response from a bot can be captured via a webhook. Because a bot run could take 3 seconds, or 30 minutes (which would time-out most any API call), a callback is done via webhook once the bot execution is completed. Alternatively, you could pass in a UID to the bot, and have the bot make a call back to your API upon completion along with the same UID to associate the return with the initial call.

Quick update, I was able to convert the Postman calls over to powershell, it appears to be working, but will need to figured some more things around how to best handle the credentials...thanks!

Glad to hear the episodes were helpful in getting remote bot execution working in your environment. As to your request about other languages, fortunately Postman does an awesome job of this for you out of box. With any of the requests opened in Postman, look for the “Code” link on the right hand side of your screen below the Send/Save buttons. This will bring up a code snippet generator that will build all the code you need to make the same request you had open in 16+ languages - using a variety of libraries within those languages. This is the best way to get a jump start on taking some sample code to build into a custom solution of your own. As to your question about the tokens/storage. The tokens are good for 20 minutes by default - though you could extend this time through your CR configuration or by using the /authentication/token endpoint to refresh a currently valid token. For the secure storage of the credentials to call the CR API, one consideration would be authenticating with an API key instead of a password to reduce the capabilities of what anyone could do should they access your value. As for where to store the values, that could be specific to the programming language you’ve selected or it could be as centralized as using something like cyberArk (or another vault system) where passwords/keys/etc can be securely and centrally stored.

@@Automationanywhere Agreed, as I continue to explore the functions of that tool, it has become more clear to me as we move forward, showing us the benefits to getting started quickly using different various language options... We're making great progress on this front so far, thanks to you!!!

Hey Sebastian - Rough guess is that ~90% of the endpoints are the same. You likely could use this exact same postman collection for the v11 API with only minor changes. You'd want to take a look at the /swagger/ endpoint for your v11 CR vs the A2019 community edition swagger to be sure though

It sounds like there many seminaries for the postman example collection Micah provided. But I agree... there is many of us that will be needing to migrate off of V11 to A2019, any tips/tricks/gotchas we'll need to know about or links to that info would be super helpful!!!

Output variables are available via CR API after we updated A2019 to 6453. It is very useful, thanks to the AA team for continuous improvements!

Glad to hear you figured it out, Alex. We're always here to help.

@@AutomationanywhereUnfortunately I could not get the botOutVariables. It would be great if you could explain better how to use this resource. Thanks in advance

Triggering a Bot through the API requires a Bot Runner license, which users don't have in Community Edition, so you wouldn't be able to use this particular endpoint in CE.

@@Automationanywhere I don't understand this licensing requirement. I am a dev in my enterprise org, so I have a creator license and access to a *development* VM. But I need to be able to run the bots remotely on that dev VM in order to develop and test using the API before handing off the code for deployment to our production environment where the bot runner licenses are used by the production team.

@@McGurble For testing of Unattended bot execution, you would need to have an Unattended Bot Runner license. The development license is designed for development and testing of bots from with in the development interface. Follow-up with your account manager if you require additional licenses.

Hi Raju. The Automation Anywhere trial version does not include the Bot Runner license required for the deployment of a bot to an unattended Bot Runner.

Hi Selorm. We recommend that you post a screenshot of your request payload on apeople.automationanywhere.com to get help with troubleshooting your code. Additionally, make sure that the User ID that you're referencing has a default device defined.

Hey Franco - you can actually trigger bots using the API the same way Micah shows in this video. At the time of recording, triggering bots in Control Room API of Community Edition was not available, but it should be working for you now. Alternatively, other triggers like file/email triggers can be tested from the bot building interface of Community Edition. If you need to test specific unattended runner capabilities, get in touch with someone from our sales team who can understand your needs and help get you the access you need for testing.

@@Automationanywhere Hi, thanks for your answer! I was able to do several tests and they worked fine! I was just a bit confused. I just have another question, can i use de Deploy Bot request in the community edition?

Hey Aphisit - check out Triggers in the Automation 360 workbench. These enable you to kick off bot runs based on different criteria.

@@Automationanywhere thank you very much.

Hey Archit - you wouldn't be able to send a file itself as the type for each value has to be one of the defined types for A2019. The only valid types are STRING, NUMBER, BOOLEAN, LIST, DICTIONARY, DATETIME. As an alternative though, you may consider triggering the bot with a string, where the string is the fully qualified path to the file up on a share drive.

Thanks for the clarification. Any recommendations if file needs to be picked from internet facing applications?

@@Archit958 I would do a custom package for that. You could feed in the file location as an input value, then call the internet facing application in the package (Java) to pass a file

@@micahman33- sounds great. Will give it a try.

Hey Roen - Check out setting up device pools and using Workload Management (WLM): docs.automationanywhere.com/bundle/enterprise-v2019/page/enterprise-cloud/topics/control-room/workload/cloud-workload-an-overview.html

Hi yuv raj, deploying a bot requires that you're using Enterprise A2019 as this functionality is not available in Community Edition. If you're already using Enterprise Edition, be sure that your role has access to deploy bots using the bot runner account in question.

Same with me. How do I get the control-room-URL. If you found it, plz let me know.

same problem here, are you find any solution?

@@yuvraj5216 1 - I Installed a lot of times many versions "on premises" not cloud, wich caused the problem, then I deleted my bot runner user and I created a new one. 2 - Don't forget to delete SQL Server Database as well, the new installation will recreate for you. 3- Assign all roles available to the new bot runner user. 4 - Don't use the server to run the bot it will not work due to some limitation of the AA tool, you need other VM or PC to run exclusively the task bot, you can use the server to start the task bot not to run. 5- Remember that I'm usign AA 2019 Enterprise, Community will never work with API. 6- If you have any bot runner previously configured make sure that you are running the correct version of this agent with "user name" and "password" correctly, I used the ouput of the follwing command "whoami" of cmd.

@@CofreDigital thank you so much

Hey dilpreet the call to trigger a bot is asynchronous - to say that the control room is not waiting for the completion of the bot execution before retuning a response to your API request. If you needed the bot to send some output, that call would need to be made from the bot itself using the REST POST action. (You may consider passing in some GUID or something so an incoming call with data from the bot can be coordinated with the outgoing call to trigger the bot). Once you've made the deployment request you can use the /v2/activity/list endpoint + the deviceID to get the status of bot executions should you need to understand if one is still running.

Try hard coding your username/password one time in the JSON body making sure that the JSON formatting for "Beautify" is selected. It may be that your password has an illegal character which would need escaping. (Double quotes and backslash both come to mind).

@@micahman33 can we use this api with community edition?

Yes, that will work!

Hi Jordi. The postman collection can be found here: www.postman.com/collections/bb5cd3541d29e192cd43

@@Automationanywhere Thank you!!

Hi @AutomationAnywhere team please respond

@Automationanywhere team please

Sounds like a permission issue! Make sure that the account you're using for your bot runner has access to execute the automation. Also, make sure that your developer account has access to trigger automation executions on another device/user's behalf

Thanks for the reccomendation @Jaimin - we'll pass it along to Micah. In the meantime, he suggested a Dev Portal tutorial on Bot Lifecycle Management (aka Repository Management) developer.automationanywhere.com/blog/bot-lifecycle-management-in-automation-360

Hey 4imble - Micah and team have dedicated Bot Runner and Bot Creator accounts because each license enables different capabilities and couldnt be applied to the same account. For example, if you were going to run a bot unattended, you couldn't use a Bot Creator license for that, it would need to be done using a Bot Runner. That said, a single Bot Runner account could/should be used across multiple bots. So a runner may be doing an HR task at 9am, but by 9:30am it may be doing a finance task or something totally unrelated. Hope that helps