Skip to main content
S
soumya-hcl1-Visitor
1-Visitor
July 16, 2018
Solved

Monitor Device Performance & Create Alerts tutorial

  • July 16, 2018
  • 1 reply
  • 2157 views

Hi All,

I am trying to go through the tutorial- Monitor Device Performance & Create Alerts using ThingWatcher and provided training/model services and simulator. Whenever I run the simulator which sends the data to the platform for calibrating the training service I get the below error.

ERROR [2018-07-16 01:50:20,730] com.thingworx.analytics.ccc.job.local.AsynchronousMemoryJobExecutor: Call to TrainingJob [20180716] failed due to:
! java.lang.IllegalArgumentException: System memory 259522560 must be at least 471859200. Please increase heap size using the --driver-memory option or spark.driver.memory in Spark configuration.

 

It clearly visible what is the issue here. I am running the ThingWorx analytics trial edition on a Windows 10 laptop with 16 GB RAM which is the minimum required for running the analytics server. The VM created as a memory of 10GB. But still it runs out of memory. If the am calculating 471859200 bytes as the required memory then it is around .471GB which is much smaller than the my VM RAM. Can anyone help.

Best answer by cmorfin

Hi

 

Thank you for clarifying, now I get what you mean.

In theory you should not need to go to the Hyper-V manager, so this VM should be mostly hidden as opposed to running docker toolbox.

 

I apologize but I overlooked one point in my previous answer.

If you are strictly following https://developer.thingworx.com/resources/guides/anomaly-detection-how-guide and using ThingWorx Analytics Trial Edition, then actually the microservices for the anomaly detection are started as java command and not as docker containers.
To resolved this memory error you then need to add more memory to the java process using the -Xmx flag . See https://www.ptc.com/en/support/article?n=CS269383 . for some more details.

 

Kind regards

Christophe

 

 

 

1 reply

C
cmorfin19-Tanzanite
19-Tanzanite
July 16, 2018

Hi

 

I am a little confused with the setup you have.

You indicated using Analytics Trial Edition on Windows 10 with 16Gb of RAM.

But you are adding the VM created has 10Gb of RAM.

Which VM is that ?

 

If using Analytics Trial on Windows 10 , you would use Docker for Windows and there is native container support, no VM  involved here.  So I do not quite understand what this VM is.

 

However you would need to ensure that the docker process has got sufficient memory. Aim for at least 5 to 6 Gb.

See https://www.ptc.com/en/support/article?n=CS252454 for the steps on how to set this up on Windows 10.

 

Hope this helps

Kind regards

Christophe

 

S
soumya-hcl1-VisitorAuthor
1-Visitor
July 16, 2018

Hi Chris,

Thanks for your answer. Docker for windows actually creates a VM called MobyLinuxVM which you can visualize in your Hyper-V manager. Attached the screenshot. Also attached the docker settings as mentioned in the article referred by you. I have already tried this setting but I still get the issue as I have already shared here. You can see that RAM added to both the VM and Docker settings is same. Internally I guess the Docker creates this VM where Analytics and ThingWorx foundation are loaded. This is a Linux VM and if you use Windows containers in docker the trial edition installation will fail as the base images of the foundation server/analytics server are Linux images. hyper-v.jpg

 

docker_settings.jpg

C
cmorfin19-TanzaniteAnswer
19-Tanzanite
July 16, 2018

Hi

 

Thank you for clarifying, now I get what you mean.

In theory you should not need to go to the Hyper-V manager, so this VM should be mostly hidden as opposed to running docker toolbox.

 

I apologize but I overlooked one point in my previous answer.

If you are strictly following https://developer.thingworx.com/resources/guides/anomaly-detection-how-guide and using ThingWorx Analytics Trial Edition, then actually the microservices for the anomaly detection are started as java command and not as docker containers.
To resolved this memory error you then need to add more memory to the java process using the -Xmx flag . See https://www.ptc.com/en/support/article?n=CS269383 . for some more details.

 

Kind regards

Christophe

 

 

 

Terms & ConditionsCookie settingsAccessibility statement