Issues on running unattended bots

[KT]

Hi,

This was my first project in BP and i used a VDI machine for my UAT environment. When i logged into the machine and ran my process inside of it (attended). It was working fine on running the site and clicking the elements. But when i try to simulate unattended mode by running BP on my local machine to trigger my UAT process. i always get the error that the site is not loaded. But when i logged into the machine. i can see the site was left open. Seems like it was able to run it but was not able to click one element to proceed to the next one. The process to click the element had a dynamic wait and sleep actions already. Just wondering why i always had issues just in unattended mode. Any areas i need to check?

Another issue i have is its prompting that the BP addin sometimes is missing on the browser. Though i can see it loading ok when ran manually.

------------------------------
K T
------------------------------

[harish.mogulluri]

Hi K T,

 It looks like you need to add wait stage after launch actioin in the object level

To understand  the root cause of issue: 
1) Check the logs to identify the stage and action bot is throwing error

2)   Make sure wait stages( dynamic wait stages) are present after launching an application or navigation from one screen over the other screen

3) If  it  failed on clicking the button which type of action are you using?( If you are using global send keys or Global send key events( Make sure Application maximiazed and activated before  clicking the button)

Make sure the extension is present in the browser instance,   Probably pin the extension in the browser. if the problem persist I will remove teh extension and readd the extension and restart the VM before running in unattended.

Which version of blue prism you are currently using ?

------------------------------
-----------------------
If I answered your query. Please mark it as the "Best Answer"

Harish Mogulluri
Lead developer
America/New_York TX
------------------------------

[John__Carter]

Hi KT  - it could be related to the way you are accessing the remote machine. If you are using RDP, you're logging into the machine and creating a desktop session, you're not viewing it as a passive observer. And when you disconnect, the desktop session is discarded.

It's possible your process fails because it's trying to do something that it can't, due to the lack of a desktop session. The annoying thing with RDP is that depending on what they are doing, some processes won't care and will run fine. Typically it's things like global send keys, global mouse click and activate window.

This document is old but still relevant - https://bpdocs.blueprism.com/z-pdf/bpe/remote-access-tools/en-us/blue-prism-6-remote-access-tools.pdf

You can 'see' this RDP effect by creating a test process that takes a screenshot: if you run the process while you are connected, the screenshot is as expected; if you run the process after you have disconnected, the screenshot is black. https://digitalexchange.blueprism.com/dx/entry/4714/solution/rmr---utility---screenshot

------------------------------
John Carter
Blue Prism
------------------------------

[KT]

Thanks Harish, i got all of the dynamic waits on it. i will look at the activate action.

Regarding the extension. i normally see the error once i terminate all browsers and run my process again. I am using version 6.10.5

------------------------------
K T
------------------------------

[KT]

Thanks John, i was reading about this and it seems about the RDP and how i do my test. I tried to run the login process from BP before running my project process and it went through fine passing the problematic area. The one that is having an issue is my process with global mouse click. 

------------------------------
K T
------------------------------

[ChrisRider]

Hi KT - I recently had an issue with attended vs unattended.  My process worked fine in attended using RDP.  When I used Control Center to run the process unattended, I was seeing problems in that it could not see some parts of the application.

My issue was caused by different screen resolutions.  When I log and run it interactively, my screen resolution was set to 1280x104.  I think when it did it using login agent, it might have been 1024x768.  My solution was to set the screen size using powershell when the Blue Prism user logs.  There is also a process available that will set the resolution (I am going to add this to me process).

This thread helped me out:  https://community.blueprism.com/discussion/how-to-set-screen-resolution-while-using-scheduler-login-agent

Your issue sounds different, but it might be worth checking out.

------------------------------
Chris Rider
lead application analyst
E. W. Scripps
cincinnati OH
------------------------------

[rsrivastava2]

Hi KT,

There can be multiple things can go wrong when you talk about Attended VS Unattended. 

It would be helpful if you can provide some screenshots of the process flow that you have created and we can provide a RCA and Solution.

Hope this helps,

------------------------------
WeitghtRPAMatey
------------------------------

[rabbanir]

Hi John,

What would the resolution to this be? I'm having the same issue where running a process when not 'watching' it through a VNC window. I got it to take a screenshot and it's as you say when I'm actually connected the screenshot is as expected but when the screenshot is taken when I'm not watching it's a black screen. I tried that link you provided but it just goes to the docs homepage.

thanks,
Raheel

[John__Carter]

Hi Raheel - use the use login agent processes to log out and log back in so that the DW has a fresh windows session. Make sure nobody connects to that machine via RDP. Then run the process. Watching via VNC should not make a difference because (unlike RDP) it is 'non-invasive'.

[rabbanir]

Hi John,

I changed the schedule to do as you say where after logging in using the login agent it logs back out and back in again before running the actual automation and this worked unattended! thanks for the work around. Still though I don't really understand why it doesn't work after the first login with the login agent, Ideally shouldn't have to use this work around and it should just work after first login.