[
https://issues.jboss.org/browse/WFLY-12167?page=com.atlassian.jira.plugin...
]
Paul Ferraro edited comment on WFLY-12167 at 6/6/19 2:45 PM:
-------------------------------------------------------------
The code in ServiceSupplier is correct. Using the new MSC API, the only way to obtain a
value provided by an ON_DEMAND service is install a service that depends on the target
service. Once the value is retrieved, the mode of the temporary service is set to
REMOVE.
However, this removal happens asynchronously. Even if there was no memory leak in MSC
this would still cause a heap spike until an available MSC thread is able to perform the
removal. To resolve that, ServiceSupplier.get() can block until the temporary service
removal is complete which would prevent any spike in heap due to a pile-up of
to-be-removed services.
was (Author: pferraro):
The code in ServiceSupplier is correct. Using the new MSC API, the only way to obtain a
value provided by an ON_DEMAND service is install a service that depends on the target
service. Once the value is retrieved, the mode of the temporary service is set to
REMOVE.
However, this removal happens asynchronously. Even if there were a memory leak in MSC
this would still cause a heap spike until some MSC thread is able to perform the removal.
To resolve that, ServiceSupplier.get() can block until the temporary service removal is
complete which would prevent any spike in heap due to a pile-up of to-be-removed
services.
Memory leak in metrics in standalone-ha configuration
-----------------------------------------------------
Key: WFLY-12167
URL:
https://issues.jboss.org/browse/WFLY-12167
Project: WildFly
Issue Type: Bug
Components: Clustering, MP Metrics, MSC
Affects Versions: 16.0.0.Final
Reporter: Bernd Stolle
Assignee: Richard Opalka
Priority: Blocker
Labels: memoryleak
Attachments: Screenshot 2019-06-06 at 11.07.00.png
When started in standalone HA configuration every request to the recently added metrics
endpoint ({{<management-if>:9990/metrics}}) lead to an increase in memory
consumption until the JVM is slowed down significantly by GC to a point where even the
requests to {{/health}} fail within a reasonable timeout (2s) and untlimately lead to
OOM.
The same issue does not occur when WildFly is started in the default standalone
configuration (non HA).
I can provide a (compressed) heap dump if required.
--
This message was sent by Atlassian Jira
(v7.12.1#712002)