For the cost of a Starbucks Latte a month, you can cover all agents across your Automate Server!
What about the agents that just stop checking in? We haven't forgotten about those either. With the Stalled Agents plugin you get a view of all agents not currently checking in. You can use the same great tools provide in the plugin to control the Automate Agent Services just like the agents that have stopped processing commands.
We also look at ScreenConnect agents and who is not checking into the ScreenConnect services. Use quick tools provided to remedy failed agent services.
This plugin is accessed via the main [View] menu and when enabled will display agents that may be stalled and not responding to commands or scripts. You can then use the tool to attempt a kill of the Automate agent service and restart of the agent services on the remote agent.
We offer 2 methods to attempt Automate agent service restarts
We do this by looking for other active agents on the same network and using on of these agents that is currently working to send RPC commands over to tell remote PC to kill and restart the Automate agent. RPC requires domain level privileges so we use the Client consoles Password tab to determine the domain and user needed to execute the RPC. You will need to have a "Domain Admin" account listed in the Password Tab of your Client console for our plugin to automate the RPC calls.
This relies solely on RPC and WMI so you will need to make sure systems can communicate over RPC on network. Make sure all systems you want to be able to manage this way get the following command run on them. Create a script if need be and apply to all agents before a failure happens.
netsh advfirewall firewall set rule group="Windows Management Instrumentation (WMI)" new enable=yes
Select a system, right click and select to restart service. The following window will appear and will actively attempt the restart. Monitor this window for success or failures.
How it all works.
The plugin queries the local Automate database to get a list of agents that have "executing" commands. We look at the tables in Automate for values runningscripts.running = 1 and commands.status = 2 and create a list of computers that have running commands counts. Agents should really only have 1 or 2 commands pending execution at any given time so to see a agent with 50 or 100 commands in a running state but a status of 2 is most likely a stalled agent.
When you select an agent to do clear and attempt a restart the plugin first update the database setting the agents executing commands to aborted then it queries the database to find a host nearby, on the same network as the failing agent. The agent is sent a powershell script that gets stored in %windir%\LTSvc\StalledAgents\ . The plugin queries the Automate database for a "Domain Admin" password to be set for the client at the client console's passwords tab. If available it uses this username and password to execute the powershell script passing the username and password to the powershell script to execute as that user. The commands needed to send RPC over to failed agent instructing a restart of the Automate service and kill process for LTSvc are included in the powershell script and can be executed manually at anytime by logging in and going to the directory listed above..
If you receive errors during a restart from the powershell commands executing, it will be displayed in the script execution box's terminal window. Allow time for the process to complete before shutting down the terminal window. Read your failures, most will be due to permissions, firewalls blocking RPC or RPC services not running on remote systems. We suggest that you take time during deployment of plugin to verify that the Domain Admin account exists and is current, RPC is allowed through all windows firewalls active on the local networks and that RPC services are available. This will allow tools like this to work quickly and effectively when large problems arise.
A fellow geek by the name of datacomm on the MSPGeek forums has created and published a ConnectWise Connect plugin called RMM+Plus that is just a awesome tool. By adding the plugin to your ConnectWise Connect manager you can use the ConnectWise Connect agent to pass commands down to the local system outside of any Automate services. This makes it quite easy to get agents restarted with little effort and no RPC issues like described above. You only need the plugins access code to get rolling with our plugin. This process also supports Mac and Linux OS types where ConnectWise Connect is installed.
For ConnectWise Connect plugin information and how to install it to LabTech see LabTechGeek Post
The free to use RMM+ tool can be installed as an plugin to ConnectWise Connect and then assigned 2 configuration variables.
The Access Key and Allowed Origin. These provide access to the Connect plugin from outside Connect (Automate).
In the Stalled Agents plugin in Automate you will find this control as part of the menu list of listed agents. It allows you to save the above Access Key and Allowed Origins settings so both plugins can communicate.
Buggy Agents Detector
There is a know bug in the Automate agent that affects some MSPs. This bug prevents the agent from allowing the scripts and commands sent to agent from working correctly. This will cause the agent to fail common tasks like updating hardware statuses and patching, yes patching! So to find these little buggers we deploy a script to test the agent to see if we get a invalid response from a basic command. If we do the agent is seen here in list. Notify Automate that you have agents in this list for then to assist you further. They will most likely put you on the bug list so you get notified when they fix the issues.