KB484051: How to prevent large dump (.mdmp) files from being generated under the Services Registration folder

Description

Remote Diagnostics is a feature that gathers customer's environment, platform, and license information to help improve Strategy products and enhance user experience. This information is collected by an

envinfo-collector.jar 

binary that is executed periodically. This periodic execution, which by default occurs every 30 minutes, is enabled whenever Strategy Services Registration is running. 
The following screenshot shows a very large

 hs_err_pid<pidnumber>.mdmp

file generated by

envinfo-collector.jar 

binary under Services Registration's directory.
 

Under certain conditions, such as very high memory utilization, the 

envinfo-collector.jar 

binary may crash and, consequently, the underlying Java Virtual Machine (JVM) will generate an 

hs_err_pid<pidnumber>.mdmp

file. Depending on the 

envinfo-collector.jar

's memory footprint, the dump file can be very large, reaching up to 3 GB in size, as shown in the above screenshot. In addition, since the 

envinfo-collector.jar 

is executed periodically, it is likely that multiple dump files are generated and further cause disk space depletion. Left untreated, the low disk space condition can lead to performance issues ranging from slow response to denial-of-service.

Workaround

Follow these steps to prevent the large 

hs_err_pid<pidnumber>.mdmp

file from being generated:
 

1. Stop the Strategy Services Registration service on all the nodes of the Topology cluster. 
   Note: The offending node, where the large

   hs_err_pid

   .mdmp

   file is located, may belong to a Topology cluster, e.g., a collection of nodes whose services can be administered through a Topology view inside Strategy Workstation.
   
2. On the offending node, open the

    mstr_services_config.yaml 

   file inside the

   yaml

   sub folder of the Services Registration Directory.
   The default path to this file in Windows is

   C:\Program Files (x86)\Services Registration\yaml\ mstr_services_config.yaml.

   
   
3. In the

   mstr_services_config.yaml 

   file, add the following entry in red:
   

   
   … 
     #This is a "dummy" check that triggers env info collection 
     #It is coexisting with REST Server 
     - 
       name: "Env-Info-Collection-Scheduler" 
       id: "Env-Info-Collection-Scheduler" 
       args: 
       - "<JavaInstallPath>/java" 
       - "-XX:-CreateMinidumpOnCrash" 
       - "-jar" 
       - "../jar/envinfo-collector.jar" 
       - "send" 
       interval: 30m 
       timeout: 15m 
   … 
   

   
   
4. Open the

   Strategy-Library-REST-Server.json

   file inside the config sub folder of the Services Registration directory.
5. In the

   Strategy-Library-REST-Server.json 

   file, add the following entry in red:
   

   …	         {	           "name": "Env-Info-Collection-Scheduler",	           "id": "Env-Info-Collection-Scheduler",	           "args":  [	             "C:\\Program Files (x86)\\Common Files\\Strategy\\JRE\\180_192\\Win64\\bin\\java",	           "-XX:-CreateMinidumpOnCrash", 	           "-jar",	           "../jar/envinfo-collector.jar",	           "send"	         ],	        "interval": "30m",	        "timeout": "15m"	     } 	…

   
   
6. Restart the Strategy Services Registration services on all the nodes.

An 

hs_err_pid<pidnumber>.log 

file is also generated together with

every hs_err_pid<pidnumber>.mdmp 

file. This log file contains basic information regarding the root cause behind the crash that generates the large dump file. The 

hs_err_pid<pidnumber>.log 

file also contains several possible solutions that may prevent the crash from occurring again. 
The following 

hs_err_pid<pidnumber>.log 

file provides an example of a root cause and potential solutions to a crash.