Skip to main content

Hello, As we noticed that Zulu Platform x64 Architecture has high memory usage.

Hence we are considering to stop the relevant service which is Automation Anywhere Bot Insight services to release the memory used by Zulu Platform x64 Architecture.

 

Before stopping Automation Anywhere Bot Insight services , we want to know that will it affect the process of history log on Control room as shown below

 

 

@Roy Ng No, it will not impact audit / history log as this log stores and uses elastic search


@Roy Ng are u sure its using high memory or high CPU ? What was RAM size and cpu configuration ?


@rbkadiyam  Thank you for your reply

Our configuration is :

RAM : 8.0 GB

CPU : Intel(R) Xeon(R) Platinum 8168 CPU @2.70Ghz

 

Sometimes our RPA job will need to visit a website with Chrome but it will occur an error  “memory not enough”.

And then, we checked our task manager which was showing around 311MB memory and 0% CPU usage used by Zulu Platform x64 Architecture(Top 1 usage)

 

Is the problem caused by our configuration?


Hi @Roy Ng ,

 

Using Excel actions especially while dealing with huge data will automatically increase the system memory usage during the execution. For such cases, using either Data Table or Database actions will be helpful. This will make the execution much faster with limited memory consumption.  

Or check the CPU and memory utilization and kill the respective process which is consuming them the most through BOT.


@Padmakumar 

 

Thank for your help. I will check this point with my teammates and keep checking the usage of RAM.


@Roy Ng 

Zulu/Java.exe process taking maximum memory on Automation 360


For the runner machines which are meeting the Product recommended hardware pre-requisites while running bots the Zulu/Java.exe takes maximum memory during run time. This usual occurs for the long running bots which more than 4 hours consistently. 

Error Message:  java.lang.OutOfMemoryError: Java heap space

Resolution/Workaround

Follow the below steps on the runner machine to reduce high memory leak:

  1. Navigate to the following folder C:\Program Files\Automation Anywhere\Bot Agent\config
  2. Create a new file collector.properties file and add the below entry

event.text.process.filetags=TriggerListener,Bot_Launcher,Node_Manager

  1. Edit nodemanager-logging.xml (at same above location) and remove the full line which is highlighted in bold  <AppenderRef ref="json"/>
  2. Ensure there are no old log files (taken as backup) at the bot agent default logging folder. Ideally just one nodemanager.log & Botlauncher.log should exist at below location and also remove all .json files (if exists) from below location:

C:\ProgramData\AutomationAnywhere\BotRunner\Logs

 

 

(or)

 

This issue can happen if one or more Bot Agent user profiles are configured with the wrong Device credentials (Login into Automation 360 >> Click on Local Device Icon >> Credentials), which will send the continuous authentication failure request to the Control Room. Please review the WebCR_RestException.log file available on the Control Room Server at the Logs location (Default Path: C:\ProgramData\AutomationAnywhere\Logs) in which you will find the below log entry continuously getting logged. (In one of the cases we had found 150 entries in 1 min of duration)

ERROR - com.automationanywhere.common.util.logutil.LogUtil - {url:https://CRURL/v1/authentication/token method:POST code:401 requestbody:null responsebody:{"code":"UM1111","details":"To be safe, we logged you out. Please login again.","message":"Your session expired"} com.automationanywhere.common.validation.DetailedException: Your session expired
 

Resolution/Workaround

Please follow the below steps to Identify the Device causing the authentication failure issue and resolve it.

       1. Please enable the traefikAccess.log as shown in the below Knowledge Base article. The traefikAccess.log file will help to identify the Device IP Address which is causing this authentication failure issue.            
           
           https://apeople.automationanywhere.com/s/article/A360-How-to-enable-traefik-access-log

       2. Please find the ClientAddr, ClientHost IP Address of the Device in the traefikAccess.log file as shown in the below sample log entry which is causing the authentication failure.

{"ClientAddr":"IPAddress:Port","ClientHost":"IPAddress","ClientPort":"Port","ClientUsername":"","DownstreamContentSize":113,"DownstreamStatus":401,"Duration":13872964900,"OriginContentSize":113,"OriginDuration":13872964900,"OriginStatus":401,"Overhead":0,"RequestAddr":"CRURL","RequestContentSize":641,"RequestCount":5111,"RequestHost":"CRURL","RequestMethod":"POST","RequestPath":"/v1/authentication/token","RequestPort":"","RequestProtocol":"HTTP/1.1","RequestScheme":"https","RetryAttempts":0

      3. Access all the Bot Agent Devices taking Remote Desktop Access which are showing in traefikAccess.log file as authentication failure and stop the Automation Anywhere Bot Agent Service. Once you stopped, then check the CPU memory used by the Zulu process on the Control Room Server, it should go down which will confirm the root cause of the issue. 

      4. Once confirmed, please start the Automation Anywhere Bot Agent Service on the device and update the Device credentials with correct information by login into the respective Bot Agent profile in the Automation 360 Control Room. 


@rbkadiyam Thank you for the workaround information. 

For the 2nd step of the first case, should I keep the only line that the file currently has?

bot.launcher.archive.enabled=false

and add below it the following?

event.text.process.filetags=TriggerListener,Bot_Launcher,Node_Manager

or the new file should only have the second one?

 

For the 3rd step,  I couldn’t find the exact line <AppenderRef ref="json"/>, but I find  <AppenderRef ref="NodeManagerJson"/>. Would it be the same?

 

Thanks for your help!


@gines.garcia  can yo provide log files to dig more ?


@rbkadiyam please find the bot runner log files attached. 

Thanks!


@rbkadiyam  @Padmakumar 

 

Thanks for your help

 

Finally, we found that problem is caused by “Windows Explorer”. It consumed around 4GB ram but task manager was showing this progress which used only 62MB ram. The 4GB memory usage is shown by Powershell with command. Therefore, we might kill the “Windows Explorer” regularly until found out the reason that why “Windows Explorer” consumed a large memory usage cumulatively.


Hi, A couple of days back this “Zulu Platform x64 Architecture” background processor was not in my system. But, now it is running while I am login in my system. How can I remove it? If is there any solution. Please guide me.


Reply