Re: build failures on jenkins
Ok, The key offender there was new SecureRandom().generateSeed(128) The test launches JMeter multiple times, and it caused lots of "generateSeed(128)" calls. I've removed excessive generateSeed(128), and I've made the initialization lazy. The stacktrace in question was is "main" #1 prio=5 os_prio=0 tid=0x7f507400c800 nid=0x33d2 runnable [0x7f507aaa] java.lang.Thread.State: RUNNABLE at java.io.FileInputStream.readBytes(Native Method) at java.io.FileInputStream.read(FileInputStream.java:255) at sun.security.provider.NativePRNG$RandomIO.readFully(NativePRNG.java:424) at sun.security.provider.NativePRNG$RandomIO.implGenerateSeed(NativePRNG.java:441) - locked <0xffb11238> (a java.lang.Object) at sun.security.provider.NativePRNG$RandomIO.access$500(NativePRNG.java:331) at sun.security.provider.NativePRNG.engineGenerateSeed(NativePRNG.java:226) at java.security.SecureRandom.generateSeed(SecureRandom.java:533) at org.apache.jorphan.util.JOrphanUtils.(JOrphanUtils.java:52) at org.apache.jmeter.util.JMeterUtils.loadJMeterProperties(JMeterUtils.java:213) at org.apache.jmeter.JMeter.initializeProperties(JMeter.java:788) at org.apache.jmeter.JMeter.start(JMeter.java:461) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.lang.reflect.Method.invoke(Method.java:498) at org.apache.jmeter.NewDriver.main(NewDriver.java:253) Vladimir
Re: build failures on jenkins
On Sat, 28 Sep 2019 at 10:34, Felix Schumacher wrote: > > Hi all, > > it seems that the builds on jenkins are failing since the last good one > seven days ago > (https://builds.apache.org/blue/organizations/jenkins/JMeter-trunk/activity). > The jobs seem to be getting killed after 30 minutes. The first change > that reached this limit was > https://builds.apache.org/blue/organizations/jenkins/JMeter-trunk/detail/JMeter-trunk/7349/pipeline/. The job has a timelimit of 30 minutes, because normally it takes a lot less. The timeout should only be increased if the tests really do need longer to complete. 7348 only took 16mins looks like that was from ee675911a8a558c0c96bedb85b9096fded95aa4b I've temporarily flagged 7348/7349 as builds to be kept forever, and increased the limit to 40 builds so people can check what changed to cause the build to take so much longer > Regards > > Felix >
