Make StandardExports work with modules and/or IBM JVMs by alin-amana · Pull Request #281 · prometheus/client_java

alin-amana · 2017-07-24T12:55:19Z

com.sun.management.UnixOperatingSystemMXBean is not available in IBM JVMs and not easily accessible in some environments, such as WildFly. Use reflection instead to collect file descriptor metrics.

…e com.sun.management.UnixOperatingSystemMXBean is not available. Use reflection instead to collect file descriptor metrics.

brian-brazil · 2017-07-24T13:02:04Z

Please see #278, which I'd like to resolve first.

alin-amana · 2017-07-24T13:08:02Z

Interesting. But I believe the instanceof operator in your code will result in a ClassNotFoundException in that case too. I wonder if there exists a programmatic approach to figuring out which interface actually declares that method and grab the method from there (whether it's com.sun, com.ibm or something else).

brian-brazil · 2017-07-24T13:12:15Z

I'd be happy to have something that works with both, but if the choice is between Oracle and IBM JVMs then given this is the hotspot module we'll go with Oracle.

alin-amana · 2017-07-24T13:17:20Z

Fair enough. All I'm saying is that the instanceof operator will throw an exception if running on an IBM JVM, so the else case is useless.

Anyway, I'll be looking into programmatically looking up the implementing interface and getting the method from there.

…lass then on all the interfaces it implements until either success or all attempts have failed

alin-amana · 2017-07-24T15:26:22Z

Gave it another try. This should cover both process_cpu_seconds_total and the file descriptor metrics.

It tries to call the method directly on the concrete class, then recursively on all implemented interfaces until it either succeeds or has failed all options.

alin-amana · 2017-07-25T13:14:49Z

Any opinion?

The code is somewhat generic, to be able to handle null return values. Since the getters we're calling all return long, it would be possible to drop the ReturnValue class and instead keep trying until we get a non-null value.

brian-brazil · 2017-07-25T13:49:44Z

As I said, I wish for #278 to be resolved first to unbreak Java 9 before adding new things.

alin-amana · 2017-07-25T14:46:54Z

Apologies for not making it clear enough, this change should also cover #278. It will traverse the tree of implemented interfaces and attempt to call the method on each of them until it either succeeds or runs out of interfaces.

Although I haven't tested it, this means it should work without issues under Java 9.

alin-amana · 2017-07-26T09:57:35Z

Friendly ping.

brian-brazil · 2017-07-26T15:29:15Z

+      mfs.add(new GaugeMetricFamily(
+          "process_max_fds", "Maximum number of open file descriptors.", maxFdCount));
+    } catch (Exception e) {
+      LOGGER.log(Level.FINE, "Could not access file descriptor metrics", e);


This is going to be spammy on non-unix systems.

I just replicated the behavior above for the CPU time metric. That, however is much less likely to log anything, particularly with the 2 known implementations.

Removed, replaced with a comment.

brian-brazil · 2017-07-26T15:33:25Z

-          unixBean.getMaxFileDescriptorCount()));
+    // There exist at least 2 UnixOperatingSystemMXBean interfaces, in com.sun.management and 
+    // com.ibm.lang.management. There are also environments (such as Wildfly) where access to these
+    // interfaces is restricted. Hence use reflection and recursively go through implemented


I don't think restricted is quite the right word, my understanding is that the way Wildfly does things breaks standard JMX usage.

Right. Looks like you need to define modules and dependencies between them, which is quite a bit of unnecessary overhead if you don't explicitly reference com.sun.something.

In the end, I dropped the reference to WildFly altogether, as it's too specific.

brian-brazil · 2017-07-26T15:34:26Z

+   */
+  static ReturnValue callMethod(Method method, Object obj) {
+    try {
+      method.setAccessible(true);


Are you sure this works on Java 9?

The call itself should work just fine. All the Java 9 AccessibleObject.setAccessible() documentation says in addition to the previous versions is that there is some logic preventing this working across module boundaries unless there is a direct module dependency and the method is public (essentially). Seeing how the two UnixOperatingSystemMXBeans are interfaces, the methods are definitely public. So in the worst case one would need to define a module dependency to make this work. Which, if necessary at all, would need to be defined for any kind of implementation to work.

To summarize, if any code will work in Java 9 then this should work too.

Do we still need this trick given the interface stuff below?

Great point. Removed, as we don't need it for our current uses.

One observation though: if we ever want to invoke a method that's private/protected (because it's only declared in the concrete implementation, not an interface) it will be necessary to add it back.

brian-brazil · 2017-07-26T15:37:10Z

+  /**
+   * Wraps a return value, similar to a Future but simplified.
+   */
+  static class ReturnValue {


This seems unnecessary, we always return a Long.

brian-brazil · 2017-07-26T15:37:19Z

+          e);
+    }
+
+    for (Class<?> clazz : method.getDeclaringClass().getInterfaces()) {


Why is this needed?

This is the "recursively visit all implemented interfaces" part.

A Method object refers to a specific class. So the first invocation of callMethod() will happen with method === SomeConcreteClass.getProcessCpuTime(), which will fail under Java 9. We then need to iterate over all interfaces implemented by SomeConcreteClass and eventually invoke callMethod() with method === UnixOperatingSystemMXBean.getProcessCpuTime(), which will then succeed.

Could you mention this in a comment?

…safe to make for our current usage.

alin-amana · 2017-07-27T08:49:55Z

Addressed all your comments. Code is now more concise, there is no additional logging and the comments have been made more relevant.

Please take another look.

brian-brazil · 2017-07-27T12:11:51Z

      try {
-        interfaceMethod = clazz.getMethod(method.getName(), method.getParameterTypes());
+        Method interfaceMethod = clazz.getMethod(method.getName(), method.getParameterTypes());
+        Long result = callLongGetter(interfaceMethod, obj);


This is recursive, do we need to worry about an infinite recursion?

No. getInterfaces() will return exactly the list of interfaces implemented/extended, in the order they are listed in the class/interface definition. An interface cannot extend itself, so it follows this is a finite sized tree.

Also, we're not really going down all branches. If the method is not declared by one of the implemented interfaces, we don't go down that branch because then that method cannot be declared by any of that interface's ancestors either. I believe there is even a constraint about a method only being defined by a single interface (i.e. you can't implement 2 interfaces with a run() method each), meaning this recursion would only ever result in a simple path traversal.

Apparently there is no constraint on implementing 2 interfaces defining the exact same method. The constraint is on 2 methods with the same parameters but different return types (obvious why that wouldn't work).

Still, the tree is finite and visiting more than a simple path is an edge case.

brian-brazil

That generally looks fine.

Have you tested this on both Java 9 and IBM's JVM?

brian-brazil · 2017-07-27T14:29:25Z

+   * might happen with method === SomeConcreteClass.publicLongGetter() and will fail if
+   * SomeConcreteClass is not public. We then recurse over all interfaces implemented by
+   * SomeConcreteClass (or extended by those interfaces and so on) until we eventually invoke
+   * callMethod() with method === SomePublicInterface.publicLongGetter(), which will then succeed.


This isn't JavaScript, just == will do

brian-brazil · 2017-07-27T14:37:03Z

+  static Long callLongGetter(Method method, Object obj) {
+    try {
+      return (Long) method.invoke(obj);
+    } catch (Exception e) {


Would IllegalAccessException be sufficient here?

Replaced with IllegalAccessException and am now throwing InvocationTargetException when the invoked method throws, thus aborting the recursion if that happens.

alin-amana · 2017-07-27T15:24:30Z

Have you tested this on both Java 9 and IBM's JVM?

Nope, not at all. But since I'm not using anything specific to a JVM implementation or other, that should mean the IBM JVM must work.

As for Java 9, the code no longer callis setAccessible(), which is what was failing. So if it is failing, it is failing for different reasons.

brian-brazil · 2017-07-27T15:27:12Z

The change looks fine, but please test it on both those platforms as we're touching on implementation-specific stuff that has been problematic before.

alin-amana · 2017-07-27T15:49:25Z

(Apologies for earlier post, wrong thread.)

OK, ran against both JVMs and the Simpleclient/Simpleclient Hotspot tests all passed.

However, FYI, I did get failures in the Spring Metrics client when running against Java 9:

$ JAVA_HOME=~/Work/jre/jre-9 ./mvnw test
[...]
Running io.prometheus.client.spring.web.MethodTimerTest
Tests run: 6, Failures: 0, Errors: 6, Skipped: 0, Time elapsed: 0.277 sec <<< FAILURE! - in io.prometheus.client.spring.web.MethodTimerTest
testOverloadedMethodName(io.prometheus.client.spring.web.MethodTimerTest)  Time elapsed: 0.186 sec  <<< ERROR!
java.lang.IllegalArgumentException: error at ::0 can't find referenced pointcut timeable
	at org.aspectj.weaver.tools.PointcutParser.parsePointcutExpression(PointcutParser.java:301)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.buildPointcutExpression(AspectJExpressionPointcut.java:217)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.checkReadyToMatch(AspectJExpressionPointcut.java:190)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.getClassFilter(AspectJExpressionPointcut.java:169)
	at org.springframework.aop.support.ComposablePointcut.<init>(ComposablePointcut.java:66)
	at org.springframework.aop.support.Pointcuts.union(Pointcuts.java:52)
	at org.springframework.aop.aspectj.annotation.InstantiationModelAwarePointcutAdvisorImpl.<init>(InstantiationModelAwarePointcutAdvisorImpl.java:92)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisor(ReflectiveAspectJAdvisorFactory.java:209)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisors(ReflectiveAspectJAdvisorFactory.java:136)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAdvisorsFromAspectInstanceFactory(AspectJProxyFactory.java:121)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAspect(AspectJProxyFactory.java:111)
	at io.prometheus.client.spring.web.MethodTimerTest.getProxy(MethodTimerTest.java:48)
	at io.prometheus.client.spring.web.MethodTimerTest.testOverloadedMethodName(MethodTimerTest.java:191)

testThrowWorks(io.prometheus.client.spring.web.MethodTimerTest)  Time elapsed: 0.005 sec  <<< ERROR!
java.lang.IllegalArgumentException: error at ::0 can't find referenced pointcut timeable
	at org.aspectj.weaver.tools.PointcutParser.parsePointcutExpression(PointcutParser.java:301)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.buildPointcutExpression(AspectJExpressionPointcut.java:217)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.checkReadyToMatch(AspectJExpressionPointcut.java:190)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.getClassFilter(AspectJExpressionPointcut.java:169)
	at org.springframework.aop.support.ComposablePointcut.<init>(ComposablePointcut.java:66)
	at org.springframework.aop.support.Pointcuts.union(Pointcuts.java:52)
	at org.springframework.aop.aspectj.annotation.InstantiationModelAwarePointcutAdvisorImpl.<init>(InstantiationModelAwarePointcutAdvisorImpl.java:92)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisor(ReflectiveAspectJAdvisorFactory.java:209)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisors(ReflectiveAspectJAdvisorFactory.java:136)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAdvisorsFromAspectInstanceFactory(AspectJProxyFactory.java:121)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAspect(AspectJProxyFactory.java:111)
	at io.prometheus.client.spring.web.MethodTimerTest.getProxy(MethodTimerTest.java:48)
	at io.prometheus.client.spring.web.MethodTimerTest.testThrowWorks(MethodTimerTest.java:121)

testSecondMethod(io.prometheus.client.spring.web.MethodTimerTest)  Time elapsed: 0.003 sec  <<< ERROR!
java.lang.IllegalArgumentException: error at ::0 can't find referenced pointcut timeable
	at org.aspectj.weaver.tools.PointcutParser.parsePointcutExpression(PointcutParser.java:301)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.buildPointcutExpression(AspectJExpressionPointcut.java:217)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.checkReadyToMatch(AspectJExpressionPointcut.java:190)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.getClassFilter(AspectJExpressionPointcut.java:169)
	at org.springframework.aop.support.ComposablePointcut.<init>(ComposablePointcut.java:66)
	at org.springframework.aop.support.Pointcuts.union(Pointcuts.java:52)
	at org.springframework.aop.aspectj.annotation.InstantiationModelAwarePointcutAdvisorImpl.<init>(InstantiationModelAwarePointcutAdvisorImpl.java:92)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisor(ReflectiveAspectJAdvisorFactory.java:209)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisors(ReflectiveAspectJAdvisorFactory.java:136)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAdvisorsFromAspectInstanceFactory(AspectJProxyFactory.java:121)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAspect(AspectJProxyFactory.java:111)
	at io.prometheus.client.spring.web.MethodTimerTest.getProxy(MethodTimerTest.java:48)
	at io.prometheus.client.spring.web.MethodTimerTest.testSecondMethod(MethodTimerTest.java:151)

timeMethod(io.prometheus.client.spring.web.MethodTimerTest)  Time elapsed: 0.003 sec  <<< ERROR!
java.lang.IllegalArgumentException: error at ::0 can't find referenced pointcut timeable
	at org.aspectj.weaver.tools.PointcutParser.parsePointcutExpression(PointcutParser.java:301)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.buildPointcutExpression(AspectJExpressionPointcut.java:217)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.checkReadyToMatch(AspectJExpressionPointcut.java:190)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.getClassFilter(AspectJExpressionPointcut.java:169)
	at org.springframework.aop.support.ComposablePointcut.<init>(ComposablePointcut.java:66)
	at org.springframework.aop.support.Pointcuts.union(Pointcuts.java:52)
	at org.springframework.aop.aspectj.annotation.InstantiationModelAwarePointcutAdvisorImpl.<init>(InstantiationModelAwarePointcutAdvisorImpl.java:92)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisor(ReflectiveAspectJAdvisorFactory.java:209)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisors(ReflectiveAspectJAdvisorFactory.java:136)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAdvisorsFromAspectInstanceFactory(AspectJProxyFactory.java:121)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAspect(AspectJProxyFactory.java:111)
	at io.prometheus.client.spring.web.MethodTimerTest.timeMethod(MethodTimerTest.java:36)

testValueParam(io.prometheus.client.spring.web.MethodTimerTest)  Time elapsed: 0.002 sec  <<< ERROR!
java.lang.IllegalArgumentException: error at ::0 can't find referenced pointcut timeable
	at org.aspectj.weaver.tools.PointcutParser.parsePointcutExpression(PointcutParser.java:301)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.buildPointcutExpression(AspectJExpressionPointcut.java:217)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.checkReadyToMatch(AspectJExpressionPointcut.java:190)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.getClassFilter(AspectJExpressionPointcut.java:169)
	at org.springframework.aop.support.ComposablePointcut.<init>(ComposablePointcut.java:66)
	at org.springframework.aop.support.Pointcuts.union(Pointcuts.java:52)
	at org.springframework.aop.aspectj.annotation.InstantiationModelAwarePointcutAdvisorImpl.<init>(InstantiationModelAwarePointcutAdvisorImpl.java:92)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisor(ReflectiveAspectJAdvisorFactory.java:209)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisors(ReflectiveAspectJAdvisorFactory.java:136)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAdvisorsFromAspectInstanceFactory(AspectJProxyFactory.java:121)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAspect(AspectJProxyFactory.java:111)
	at io.prometheus.client.spring.web.MethodTimerTest.getProxy(MethodTimerTest.java:48)
	at io.prometheus.client.spring.web.MethodTimerTest.testValueParam(MethodTimerTest.java:55)

testHelpParam(io.prometheus.client.spring.web.MethodTimerTest)  Time elapsed: 0.004 sec  <<< ERROR!
java.lang.IllegalArgumentException: error at ::0 can't find referenced pointcut timeable
	at org.aspectj.weaver.tools.PointcutParser.parsePointcutExpression(PointcutParser.java:301)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.buildPointcutExpression(AspectJExpressionPointcut.java:217)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.checkReadyToMatch(AspectJExpressionPointcut.java:190)
	at org.springframework.aop.aspectj.AspectJExpressionPointcut.getClassFilter(AspectJExpressionPointcut.java:169)
	at org.springframework.aop.support.ComposablePointcut.<init>(ComposablePointcut.java:66)
	at org.springframework.aop.support.Pointcuts.union(Pointcuts.java:52)
	at org.springframework.aop.aspectj.annotation.InstantiationModelAwarePointcutAdvisorImpl.<init>(InstantiationModelAwarePointcutAdvisorImpl.java:92)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisor(ReflectiveAspectJAdvisorFactory.java:209)
	at org.springframework.aop.aspectj.annotation.ReflectiveAspectJAdvisorFactory.getAdvisors(ReflectiveAspectJAdvisorFactory.java:136)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAdvisorsFromAspectInstanceFactory(AspectJProxyFactory.java:121)
	at org.springframework.aop.aspectj.annotation.AspectJProxyFactory.addAspect(AspectJProxyFactory.java:111)
	at io.prometheus.client.spring.web.MethodTimerTest.getProxy(MethodTimerTest.java:48)
	at io.prometheus.client.spring.web.MethodTimerTest.testHelpParam(MethodTimerTest.java:85)


Results :

Tests in error: 
  MethodTimerTest.testOverloadedMethodName:191->getProxy:48 » IllegalArgument er...
  MethodTimerTest.testThrowWorks:121->getProxy:48 » IllegalArgument error at ::0...
  MethodTimerTest.testSecondMethod:151->getProxy:48 » IllegalArgument error at :...
  MethodTimerTest.timeMethod:36 » IllegalArgument error at ::0 can't find refere...
  MethodTimerTest.testValueParam:55->getProxy:48 » IllegalArgument error at ::0 ...
  MethodTimerTest.testHelpParam:85->getProxy:48 » IllegalArgument error at ::0 c...

Tests run: 6, Failures: 0, Errors: 6, Skipped: 0

[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary:
[INFO] 
[INFO] Prometheus Java Suite .............................. SUCCESS [  0.181 s]
[INFO] Prometheus Java Simpleclient ....................... SUCCESS [  4.100 s]
[INFO] Prometheus Java Simpleclient Common ................ SUCCESS [  0.377 s]
[INFO] Prometheus Java Simpleclient Caffeine .............. SUCCESS [  0.652 s]
[INFO] Prometheus Java Simpleclient Dropwizard ............ SUCCESS [  0.356 s]
[INFO] Prometheus Java Simpleclient Graphite Bridge ....... SUCCESS [  0.333 s]
[INFO] Prometheus Java Simpleclient Hibernate ............. SUCCESS [  0.985 s]
[INFO] Prometheus Java Simpleclient guava ................. SUCCESS [  0.499 s]
[INFO] Prometheus Java Simpleclient Servlet ............... SUCCESS [  0.901 s]
[INFO] Prometheus Java Simpleclient Hotspot ............... SUCCESS [  0.601 s]
[INFO] Prometheus Java Simpleclient Httpserver ............ SUCCESS [  0.437 s]
[INFO] Prometheus Java Simpleclient log4j ................. SUCCESS [  0.358 s]
[INFO] Prometheus Java Simpleclient log4j2 ................ SUCCESS [  0.453 s]
[INFO] Prometheus Java Simpleclient logback ............... SUCCESS [  0.386 s]
[INFO] Prometheus Java Simpleclient Pushgateway ........... SUCCESS [ 14.909 s]
[INFO] Prometheus Java Simpleclient Spring Metrics ........ FAILURE [  0.630 s]
[INFO] Prometheus Java Simpleclient Spring Boot Metric .... SKIPPED
[INFO] Prometheus Java Simpleclient Jetty ................. SKIPPED
[INFO] Prometheus Java Simpleclient Jetty JDK 8 ........... SKIPPED
[INFO] Prometheus Java Simpleclient Vert.x ................ SKIPPED
[INFO] Prometheus Java Client Benchmarks .................. SKIPPED
[INFO] ------------------------------------------------------------------------
[INFO] BUILD FAILURE
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 26.729 s
[INFO] Finished at: 2017-07-27T17:42:49+02:00
[INFO] Final Memory: 18M/60M
[INFO] ------------------------------------------------------------------------
[ERROR] Failed to execute goal org.apache.maven.plugins:maven-surefire-plugin:2.15:test (default-test) on project simpleclient_spring_web: There are test failures.
[ERROR] 
[ERROR] Please refer to /home/alin/Work/prometheus_client_java/simpleclient_spring_web/target/surefire-reports for the individual test results.
[ERROR] -> [Help 1]
[ERROR] 
[ERROR] To see the full stack trace of the errors, re-run Maven with the -e switch.
[ERROR] Re-run Maven using the -X switch to enable full debug logging.
[ERROR] 
[ERROR] For more information about the errors and possible solutions, please read the following articles:
[ERROR] [Help 1] http://cwiki.apache.org/confluence/display/MAVEN/MojoFailureException
[ERROR] 
[ERROR] After correcting the problems, you can resume the build with the command
[ERROR]   mvn <goals> -rf :simpleclient_spring_web

brian-brazil · 2017-07-27T15:52:27Z

Spring tends to do complicated stuff, so I wouldn't worry about that in this context.

Considering we're swallowing errors can you run something simple like https://github.com/RobustPerception/java_examples/tree/master/java_simple and verify the process stats are there?

alin-amana · 2017-07-27T15:56:46Z

I have been running the code against my own WildFly deployed project and the stats are all there (CPU and file descriptor metrics in particular).

brian-brazil · 2017-07-27T16:36:40Z

That gives confidence, but WildFly messes with JMX. Can you try it with something simple like the above?

alin-amana · 2017-07-28T07:05:04Z

Tested. Works. On all 8, 9 and IBM.

brian-brazil · 2017-07-28T07:25:45Z

Great, thanks!

Make StandardExports work in restricted environments or IBM JVMs wher…

d63b14e

…e com.sun.management.UnixOperatingSystemMXBean is not available. Use reflection instead to collect file descriptor metrics.

Use reflection to recursively try invoking a method on the concrete c…

1ce033b

…lass then on all the interfaces it implements until either success or all attempts have failed

Cosmetic changes.

fc92b7e

brian-brazil reviewed Jul 26, 2017

View reviewed changes

alin-amana added 2 commits July 27, 2017 10:35

Review: simplify the code, clean up comments.

d668067

Spelled out the assumptions in the callLongGetter() logic, which are …

7ee0c2b

…safe to make for our current usage.

alin-amana changed the title ~~Make StandardExports work in restricted environments or IBM JVMs~~ Make StandardExports work with modules and/or IBM JVMs Jul 27, 2017

brian-brazil reviewed Jul 27, 2017

View reviewed changes

Review: minor cleanup of code and comments.

67720bb

brian-brazil reviewed Jul 27, 2017

View reviewed changes

Improve handling of exceptions thrown by invoked getters.

71098fb

brian-brazil merged commit ce76a49 into prometheus:master Jul 28, 2017

brian-brazil mentioned this pull request Jul 28, 2017

Support direct access to com.sun.management.OperatingSystemMXBean #278

Closed

MrEasy mentioned this pull request Apr 25, 2019

Package com.sun.management used directly #478

Closed

Uh oh!

Conversation

alin-amana commented Jul 24, 2017

Uh oh!

brian-brazil commented Jul 24, 2017

Uh oh!

alin-amana commented Jul 24, 2017

Uh oh!

brian-brazil commented Jul 24, 2017

Uh oh!

alin-amana commented Jul 24, 2017

Uh oh!

alin-amana commented Jul 24, 2017

Uh oh!

alin-amana commented Jul 25, 2017

Uh oh!

brian-brazil commented Jul 25, 2017

Uh oh!

alin-amana commented Jul 25, 2017

Uh oh!

alin-amana commented Jul 26, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

alin-amana commented Jul 27, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

brian-brazil left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

alin-amana commented Jul 27, 2017

Uh oh!

brian-brazil commented Jul 27, 2017

Uh oh!

alin-amana commented Jul 27, 2017

Uh oh!

brian-brazil commented Jul 27, 2017

Uh oh!

alin-amana commented Jul 27, 2017

Uh oh!

brian-brazil commented Jul 27, 2017