This article provides a brief checklist of items to run through when your server is frozen, or goes down entirely. Collecting this information before opening a support ticket is a significant step towards helping us to provide you with a quick resolution.
Resolution
To help our Support Team serve you in the most effective way, please provide the following details upon opening a support ticket:
- portal-ext.properties
- All relevant logs/stacktraces
- List of patches installed in environment
- JVM settings
- Look back to changes made in the portal before the server stopped working. Can you determine what change to the environment introduced the instability?
- While there are many reasons a server can go down, the most common are:
- CPU spiking or hanging
In this case, please provide thread dumps (5-10 dumps at 20-30 seconds apart).
- Memory
If you observe out of memory exceptions or memory leaks, please provide a heap dump. To get it to us, open a ticket as soon as possible and our engineers will assist you in FTP'ing the dumps to Liferay Support.
- CPU spiking or hanging
Providing these items for us on a ticket gives us a big headstart in pinpointing the cause of the issue. Of course, in the event of a product defect, we will be providing you a long term fix and/or a short term workaround as soon as we can.
Additional Information
Many production-down situations can be prevented through the use of an adequate non-production environment. Testing each deployment in a pre-production is a critical step to avoid production-down situations.