Skip to main content
Solved

Error creating a RDP session for half a day

  • Oct 12, 2026 7:46 AM UTC
  • 9 replies
  • 128 views
Forum|alt.badge.img

Hi everyone,

We have bots in production that have run for several months without any particular deployment problem 

On this week-end for half a day we had a systematic deployment error “An unexpected problem occurred” on both production VMs that we can’t explain. There is no information about updates or maintenance on AWS side. Moreover on a third VM not linked with that project, bots ran correctly.

The error solved itself alone and at the same time bots started running correctly on both VMs.

It was during the week-end and to my knowledge nobody did connection, admin action or maintenance on the VMs

Do you have an idea on what could be the cause of this problem?

 

Best answer by raegan.smith

​@Augustin 

I dug into this a little more and it might be helpful if you come across this issue again. 

I captured the node manager logs before and after the failed run - you have to be quick to grab them as they may overwrite depending on your set-up. I used the new AI Chat in Documentation and had it review excerpts of the logs for me to identify the root cause. The return response did give us a few directions to go in/ steps to take before submitting a support ticket with Automation Anywhere. Additional resources were provided as well. 

Our root cause lies with an issue on the Windows/RDS side of things. 

Hope this helps! 

View original

    9 replies

    Forum|alt.badge.img
    • Augustin
    • Author
    • Flight Specialist | Tier 4
    • Oct 12, 2026 12:34 PM UTC

    It seems that the bot agent version upgraded from 23.0.32 to 23.0.35 in the meantime but it seems strange that it took more than 10 hours

      Aaron.Gleason
      Automation Anywhere Team
      Forum|alt.badge.img+6
      • Aaron.Gleason
      • Automation Anywhere Team
      • Oct 12, 2026 1:57 PM UTC

      ​@Augustin Ah, the “what changed” conundrum. This is never easy or pleasant. You have to find out what changed. Chrome updates. Windows updates. Updates on the target system. RDP updates. Something changed, and it’s up to you to figure out what and revert it or fix the bots for the “new norm”.

      • Karissa M
      • Cadet | Tier 2
      • Oct 13, 2026 3:22 PM UTC

      Is this a new-ish server? I only ask because you said “several months”. 

      When you first set up an RDS server, Microsoft provides a 120-day licensing grace period. During this time, a temporary multi-user license allows users to connect without requiring a permanent Client Access License (CAL).

      Once those 120 days (roughly 4 months) expire, the server will block new remote connections until you install valid RDS CALs and configure a licensing server.

      Forum|alt.badge.img
      • Augustin
      • Author
      • Flight Specialist | Tier 4
      • Oct 13, 2026 4:04 PM UTC

      Is this a new-ish server? I only ask because you said “several months”. 

      When you first set up an RDS server, Microsoft provides a 120-day licensing grace period. During this time, a temporary multi-user license allows users to connect without requiring a permanent Client Access License (CAL).

      Once those 120 days (roughly 4 months) expire, the server will block new remote connections until you install valid RDS CALs and configure a licensing server.

       

      Hi Karissa,

      Thanks for your answer and the advice. One of the VM has been running for more than a year, the other one had just been provisioned a month ago.

      I don’t think that this is the cause of the error in that particular case

      Regards,

        raegan.smith
        Cadet | Tier 2
        Forum|alt.badge.img
        • raegan.smith
        • Cadet | Tier 2
        • Oct 13, 2026 5:51 PM UTC

        Hello ​@Augustin - my team has been facing this issue sporadically throughout 2026. We’re also on the hunt to identify this problem. Some days we experience 3+ RDP errors and other times, we go weeks without one. We operate on a single server and have 12+ bot accounts that run. No notifications are sent to notify us of the error, so we have to constantly watch the historical tab in control room.

        Are you successful in running the process with a manual trigger after you receive the RDP error or do you receive it no matter the trigger type (schedule, trigger, or manual?)

          Cadet | Tier 2
          Forum|alt.badge.img+1
          • Ken 3236
          • Cadet | Tier 2
          • Oct 13, 2026 6:22 PM UTC

          ​@Augustin Ah, the “what changed” conundrum. This is never easy or pleasant. You have to find out what changed. Chrome updates. Windows updates. Updates on the target system. RDP updates. Something changed, and it’s up to you to figure out what and revert it or fix the bots for the “new norm”.

          ​@Aaron.Gleason ‘Something changed’ can be true, and with automatic updates from the various vendors it’s nearly impossible to figure it out and would take too many resources.  We’ve resorted to simply kicking off the process again and it runs fine…...if something needed reverted or fixed you would think we get the same error, and we don’t.  

          Our RDP error only appears when we the process is being initiated from the schedule.

          Ken
          Forum|alt.badge.img
          • Augustin
          • Author
          • Flight Specialist | Tier 4
          • Oct 19, 2026 11:07 AM UTC

          Hello ​@Augustin - my team has been facing this issue sporadically throughout 2026. We’re also on the hunt to identify this problem. Some days we experience 3+ RDP errors and other times, we go weeks without one. We operate on a single server and have 12+ bot accounts that run. No notifications are sent to notify us of the error, so we have to constantly watch the historical tab in control room.

          Are you successful in running the process with a manual trigger after you receive the RDP error or do you receive it no matter the trigger type (schedule, trigger, or manual?)

          Hi and thanks for your answer.

          We have not faced the error again since. It seems to have solved itself during the week-end.

          However there were notifications of the error though e-mails.

            raegan.smith
            Cadet | Tier 2
            Forum|alt.badge.img
            • raegan.smith
            • Cadet | Tier 2
            • Answer
            • Oct 19, 2026 7:03 PM UTC

            ​@Augustin 

            I dug into this a little more and it might be helpful if you come across this issue again. 

            I captured the node manager logs before and after the failed run - you have to be quick to grab them as they may overwrite depending on your set-up. I used the new AI Chat in Documentation and had it review excerpts of the logs for me to identify the root cause. The return response did give us a few directions to go in/ steps to take before submitting a support ticket with Automation Anywhere. Additional resources were provided as well. 

            Our root cause lies with an issue on the Windows/RDS side of things. 

            Hope this helps! 

            Forum|alt.badge.img
            • Augustin
            • Author
            • Flight Specialist | Tier 4
            • Oct 20, 2026 6:51 AM UTC

            ​@Augustin 
            I captured the node manager logs before and after the failed run - you have to be quick to grab them as they may overwrite depending on your set-up. 

            This is so true!

            By default the logging settings are set to 10 files totalling 10MB, that only keep data for a day.

            I don’t know why it is not 100 files for 100MB by default, it is from now on the first thing I do when I provision a new VM

            And I don’t know why text logging would take so much space

            Terms & ConditionsAccessibility statement