Loading

What Information to Collect Raising for a CloudHub Case

Udgivelsesdato: Mar 2, 2024
Opgave

GOAL

CloudHub applications do not preserve certain pieces of very important information when restarted. The examples are:
  • A heap dump;
  • A mule Runtime mule_ee.log log;
  • Contents of application data in MULE_HOME/.mule folder;
  • Contents of /tmp directory;
  • A thread dump;
  • Other information that might help to identify a root cause.
For example, the mule_ee log can answer a question of why an application has not deployed successfully. Or, for example, what caused the runtime restart on the worker. Was it a memory-related issue, e.g., out-of-heap memory exception; or an SSL handshake issue; or something else?
Therefore, it is very important when reporting support cases, to collect the information which might be lost prior to restarting an application.  
Trin
Since the CloudHub Release 1.92.2 the feature to download the mule_ee log file of the latest deployment was introduced. Therefore, when raising a case, before restarting the application, it is recommended to follow the steps below:
  1. Do not restart the application to ensure data is preserved and collected;
  2. Download the following logs (from every application worker):
    1. mule_ee log;
    2. Diagnostics log. Make sure several diagnostic logs, i.e., 3-5 from the same deployment for each application worker, are downloaded with some interval, e.g., 30 seconds or more. Each diagnostics log represents the current Java thread dump every time it is downloaded from the UI;
    3. Application log.
  3. Check the mule_ee.log for an error there:
    1. If an error is related to any memory-related error like java.lang.OutOfMemoryError or an application is restarted being unresponsive, avoid restarting the application as MuleSoft Support most likely need to collect additional information from the application worker. Proceed to step 5.
    2. For any other error proceed to step 4.
  4. Restart the application if needed.
  5. Raise a support case with the severity as per Choosing the right Severity when reporting cases. Do not forget to provide the following:
    1. Choose the project as: Anypoint Platorm -> CloudHub -> Applications;
    2. Exact application name as deployed to CloudHub;
    3. Make sure "MuleSoft May Access My CloudHub Workers" is chosen;
    4. Add your initial analysis to the case description.
    5. Attach the information gathered.
  6. MuleSoft Support will collect the full diagnostics archive from the application worker and let them know to the application can be restarted if needed.
In some cases, when there is an immediate business impact in the production environment it is understandable that waiting for MuleSoft Support to collect data from the app workers for too long may lead to negative outcomes. However, please consider the following:
  • For production outage severity 1 cases the response time is strict;
  • Once MuleSoft Support responds, it takes 10-15 minutes approx. to collect the full troubleshooting package from application workers and let the customer know the application can be restarted;
  • if all data is not collected before the application is restarted MuleSoft Support might need to wait until the issue reoccurs again and then collect the full troubleshooting data. In this case, customers are advised to follow this procedure from scratch but instead of raising an additional case update the already raised one.
Following the above steps may drastically reduce the time required to troubleshoot the issue and provide the root cause analysis and further steps to resolve it.

USEFUL LINKS

MuleSoft Recommended Support Process

 

 

Vidensartikelnummer

001116254

 
Indlæser
Salesforce Help | Article