Code changed in jenkins
User: Gregory Boissinot
Path:
src/main/java/org/jenkinsci/plugins/fstrigger/triggers/FileNameTrigger.java
http://jenkins-ci.org/commit/fstrigger-plugin/132a6a6a2631adafd4de12173ceafa5040fdc82f
Log:
Fix JENKINS-12924

This is with version 0.29:

Polling started on Mar 7, 2012 4:16:51 PM
Polling for the job ...
Searching a node to run the polling for the label '...'.
Polling remotely on ...
Checking one file: '...'.
Slave(s) were offline at startup. Waiting for next schedule to check if there are modifications.
Checking one file: '...'.
The file didn't exist for the previous polling and now it exists.

Polling complete. Took 0.7 sec.
Changes found. Scheduling a build.

Code changed in jenkins
User: Gregory Boissinot
Path:
pom.xml
src/main/java/org/jenkinsci/plugins/fstrigger/triggers/FileNameTrigger.java
http://jenkins-ci.org/commit/fstrigger-plugin/9ff1c5f813efa172812985bfcebf68bd1cb4ac11
Log:
Fix JENKINS-12924

Happens also on latest Jenkins: 1.504, with FSTrigger: 0.35

@Ghenadie Do you have the ability to isolate the issue?
I'd to reproduce it in my environment.
Thanks

Hi,

The issue was isolated.
It happens every time the Master is restarted (Jenkins master is started under Tomcat) and master has a number of slaves connected through SSH on different machines: all machines (master+slave) are CentOS.
I have configured few jobs on each master, jobs are tied to specific slaves. On some slaves, jobs are triggered to run as described in issue once the master is restarted. Not on all slaves, which is little weird.