Skip to:

  1. Skip to Jira Navigation
  2. Skip to Side Navigation
  3. Skip to Main Content

possible issue in JCIFS Monitor - contiously increase of threads - finally heap dump

Description

We add JCIFS Monitoring at 21 Nov 3:45pm and do a OpenNMS restart.
We see a continously increase of java threads since this restart.

At Nov 23 1:16pm OpenNMS crashed with "java.lang.OutOfMemoryError" and we get a heap dump.

 

Only two servers become the CIFS pollers with different services:

 

 

Environment

Java Version: 1.8.0_222 Private Build Java Virtual Machine: 25.222-b10 Private Build Operating System: Linux 4.15.0-58-generic (amd64) OSGi Container: Apache Karaf 4.2.6 Servlet Container: jetty/9.4.18.v20190429 (Servlet Spec 3.1)

Acceptance / Success Criteria

None

Attachments

8

Lucidchart Diagrams

Activity

Show:

Christian Pape December 11, 2019 at 8:50 AM

Merged.

Martin Lärcher December 4, 2019 at 8:12 AM

Test new JCIFS Monitor and make a thread dump after one hour running OpenNMS.
No blocked threads listed in thread dump

upload thread dump threaddump_09-10-01.log

Christian Pape December 3, 2019 at 5:53 PM

I updated the branch again. It looks like timeouts were not applied correctly. Can you give it another try?

Martin Lärcher December 3, 2019 at 11:50 AM

Tested with newer jcifs  version 1.3.19 and see also a growing number of threads.
The growing is lesser but it also exists

uploaded two new thread dumps:

initial: threaddump_07-45-24.log
after 4 hours: threaddump_12-40-01.log

Christian Pape December 2, 2019 at 11:30 PM

Seems to work with an updated version of JCIFS.

Fixed

Details

Assignee

Reporter

Components

Sprint

Fix versions

Affects versions

Priority

PagerDuty

Created November 25, 2019 at 8:10 AM
Updated December 11, 2019 at 1:21 PM
Resolved December 11, 2019 at 8:50 AM