Classloader and removal of native libraries

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Classloader and removal of native libraries

Conrad Crampton

Hi,

First time posting here so ‘hi’.

I have been using Flink (1.31 now) for a couple of months now and loving it. My deployment is to JobManager running as a long running session on Yarn.

I have a problem where I have a Flink streaming job that involves loading native libraries as part of one of the mappers (inheriting from RichMapFunction) loading (in the open method) a previously trained machine learning model (using Deeplearning4j). The problem lies with when loading the model, it also loads some native libraries using javacpp Loader class (which from looking at the source code determines a location for native libraries and from a hierarchy of availability of a System property, the users home dir (with .javacpp) or temp dir).

Anyway the actual problem lies is if an exception is thrown in the Flink job, the jobmanager tries to restart it, however it would appear that when it failed in the first place, references to the objects and therefore the classes aren’t released by the classloader as I get an error

 

java.lang.ExceptionInInitializerError

        at org.nd4j.linalg.cpu.nativecpu.ops.NativeOpExecutioner.<init>(NativeOpExecutioner.java:43)

        at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)

        at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)

        at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)

        at java.lang.reflect.Constructor.newInstance(Constructor.java:422)

        at java.lang.Class.newInstance(Class.java:442)

        at org.nd4j.linalg.factory.Nd4j.initWithBackend(Nd4j.java:5784)

        at org.nd4j.linalg.factory.Nd4j.initContext(Nd4j.java:5694)

        at org.nd4j.linalg.factory.Nd4j.<clinit>(Nd4j.java:184)

        at org.deeplearning4j.util.ModelSerializer.restoreMultiLayerNetwork(ModelSerializer.java:188)

        at org.deeplearning4j.util.ModelSerializer.restoreMultiLayerNetwork(ModelSerializer.java:138)

        at com.secdata.gi.detect.map.DGAClassifierMapFunction.open(DGAClassifierMapFunction.java:54)

        at org.apache.flink.api.common.functions.util.FunctionUtils.openFunction(FunctionUtils.java:36)

        at org.apache.flink.streaming.api.operators.AbstractUdfStreamOperator.open(AbstractUdfStreamOperator.java:111)

        at org.apache.flink.streaming.runtime.tasks.StreamTask.openAllOperators(StreamTask.java:376)

        at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:253)

        at org.apache.flink.runtime.taskmanager.Task.run(Task.java:702)

        at java.lang.Thread.run(Thread.java:745)

Caused by: java.lang.RuntimeException: ND4J is probably missing dependencies. For more information, please refer to: http://nd4j.org/getstarted.html

        at org.nd4j.nativeblas.NativeOpsHolder.<init>(NativeOpsHolder.java:51)

        at org.nd4j.nativeblas.NativeOpsHolder.<clinit>(NativeOpsHolder.java:19)

        ... 18 more

Caused by: java.lang.UnsatisfiedLinkError: no jnind4jcpu in java.library.path

        at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1865)

        at java.lang.Runtime.loadLibrary0(Runtime.java:870)

        at java.lang.System.loadLibrary(System.java:1122)

        at org.bytedeco.javacpp.Loader.loadLibrary(Loader.java:963)

        at org.bytedeco.javacpp.Loader.load(Loader.java:764)

        at org.bytedeco.javacpp.Loader.load(Loader.java:671)

        at org.nd4j.nativeblas.Nd4jCpu.<clinit>(Nd4jCpu.java:10)

        at java.lang.Class.forName0(Native Method)

        at java.lang.Class.forName(Class.java:348)

        at org.bytedeco.javacpp.Loader.load(Loader.java:726)

        at org.bytedeco.javacpp.Loader.load(Loader.java:671)

        at org.nd4j.nativeblas.Nd4jCpu$NativeOps.<clinit>(Nd4jCpu.java:62)

        at java.lang.Class.forName0(Native Method)

        at java.lang.Class.forName(Class.java:264)

        at org.nd4j.nativeblas.NativeOpsHolder.<init>(NativeOpsHolder.java:29)

        ... 19 more

Caused by: java.lang.UnsatisfiedLinkError: Native Library /home/yarn/.javacpp/cache/blob_a87e49f9a475a9dc4296f6afbc3ae171dc821d19/org/nd4j/nativeblas/linux-x86_64/libjnind4jcpu.so already loaded in another classloader

        at java.lang.ClassLoader.loadLibrary0(ClassLoader.java:1903)

        at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1822)

        at java.lang.Runtime.load0(Runtime.java:809)

        at java.lang.System.load(System.java:1086)

        at org.bytedeco.javacpp.Loader.loadLibrary(Loader.java:943)

 

I’m no expert in class loading at all but have tried a number of ways to overcome this based on the docs and dynamic classloading

 

1.       Putting the jars that contain the native libraries in the flink/lib directory so they are always loaded on JobManager startup and removing them from the deployed flink job

2.       Some code in the open / close methods of my mapper to try and dispose of the native library
NativeLibrary lib = NativeLibrary.getInstance(Loader.getCacheDir() + "/cache/nd4j-native-0.8.0-" + Loader.getPlatform() +  ".jar/org/nd4j/nativeblas/" + Loader.getPlatform() + "/libjnind4jcpu.so");
this.model = null;

if (null != lib) {
    System.out.println("found lib - removing");
    lib.dispose();
    System.gc();
}
This sort of works in finding the library (on my mac running locally) but as from the stacktrace above, on deployment to yarn some random value is placed after the /cache dir which I don’t know how to get a handle on to be able to construc the correct library location (and just using the library name as the jni docs suggest for NativeLibrary.getInstance fails to find the library)

 

Neither of the above approaches work, so when my job fails I can’t cancel and resubmit as it just fails with the same stack trace. The only way I can get it to run again is the cancel all other jobs running on JobManager, killing the JM yarn session, creating a new Yarn session JM, resubmitting flink job – which is a real pain. Ideally I would like to stop the exception in the first place, but as I can’t figure out how to get logging appearing in my yarn logs either (for debug) I’m at a bit of a loss!

 

Any pointers, suggestions please??

 

Many thanks

Conrad

 



SecureData, combating cyber threats


The information contained in this message or any of its attachments may be privileged and confidential and intended for the exclusive use of the intended recipient. If you are not the intended recipient any disclosure, reproduction, distribution or other dissemination or use of this communications is strictly prohibited. The views expressed in this email are those of the individual and not necessarily of SecureData Europe Ltd. Any prices quoted are only valid if followed up by a formal written quote.

SecureData Europe Limited. Registered in England & Wales 04365896. Registered Address: SecureData House, Hermitage Court, Hermitage Lane, Maidstone, Kent, ME16 9NT

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Classloader and removal of native libraries

Aljoscha Krettek
Hi Conrad,


The code section has a good description of what is going on. We're using NativeLibraryLoader [1], which comes with RocksDB  to try and load the native library from a different temporary location if loading it the normal way fails. (The relocation of the lib to a temp dir is in the NativeLibraryLoader, not on our side. We're just providing a temp path for NativeLibraryLoader to work with.)

Hope that helps,
Aljoscha


On 10. Aug 2017, at 15:36, Conrad Crampton <[hidden email]> wrote:

Hi,
First time posting here so ‘hi’.
I have been using Flink (1.31 now) for a couple of months now and loving it. My deployment is to JobManager running as a long running session on Yarn.
I have a problem where I have a Flink streaming job that involves loading native libraries as part of one of the mappers (inheriting from RichMapFunction) loading (in the open method) a previously trained machine learning model (using Deeplearning4j). The problem lies with when loading the model, it also loads some native libraries using javacpp Loader class (which from looking at the source code determines a location for native libraries and from a hierarchy of availability of a System property, the users home dir (with .javacpp) or temp dir).
Anyway the actual problem lies is if an exception is thrown in the Flink job, the jobmanager tries to restart it, however it would appear that when it failed in the first place, references to the objects and therefore the classes aren’t released by the classloader as I get an error
 
java.lang.ExceptionInInitializerError
        at org.nd4j.linalg.cpu.nativecpu.ops.NativeOpExecutioner.<init>(NativeOpExecutioner.java:43)
        at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
        at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
        at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
        at java.lang.reflect.Constructor.newInstance(Constructor.java:422)
        at java.lang.Class.newInstance(Class.java:442)
        at org.nd4j.linalg.factory.Nd4j.initWithBackend(Nd4j.java:5784)
        at org.nd4j.linalg.factory.Nd4j.initContext(Nd4j.java:5694)
        at org.nd4j.linalg.factory.Nd4j.<clinit>(Nd4j.java:184)
        at org.deeplearning4j.util.ModelSerializer.restoreMultiLayerNetwork(ModelSerializer.java:188)
        at org.deeplearning4j.util.ModelSerializer.restoreMultiLayerNetwork(ModelSerializer.java:138)
        at com.secdata.gi.detect.map.DGAClassifierMapFunction.open(DGAClassifierMapFunction.java:54)
        at org.apache.flink.api.common.functions.util.FunctionUtils.openFunction(FunctionUtils.java:36)
        at org.apache.flink.streaming.api.operators.AbstractUdfStreamOperator.open(AbstractUdfStreamOperator.java:111)
        at org.apache.flink.streaming.runtime.tasks.StreamTask.openAllOperators(StreamTask.java:376)
        at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:253)
        at org.apache.flink.runtime.taskmanager.Task.run(Task.java:702)
        at java.lang.Thread.run(Thread.java:745)
Caused by: java.lang.RuntimeException: ND4J is probably missing dependencies. For more information, please refer to: http://nd4j.org/getstarted.html
        at org.nd4j.nativeblas.NativeOpsHolder.<init>(NativeOpsHolder.java:51)
        at org.nd4j.nativeblas.NativeOpsHolder.<clinit>(NativeOpsHolder.java:19)
        ... 18 more
Caused by: java.lang.UnsatisfiedLinkError: no jnind4jcpu in java.library.path
        at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1865)
        at java.lang.Runtime.loadLibrary0(Runtime.java:870)
        at java.lang.System.loadLibrary(System.java:1122)
        at org.bytedeco.javacpp.Loader.loadLibrary(Loader.java:963)
        at org.bytedeco.javacpp.Loader.load(Loader.java:764)
        at org.bytedeco.javacpp.Loader.load(Loader.java:671)
        at org.nd4j.nativeblas.Nd4jCpu.<clinit>(Nd4jCpu.java:10)
        at java.lang.Class.forName0(Native Method)
        at java.lang.Class.forName(Class.java:348)
        at org.bytedeco.javacpp.Loader.load(Loader.java:726)
        at org.bytedeco.javacpp.Loader.load(Loader.java:671)
        at org.nd4j.nativeblas.Nd4jCpu$NativeOps.<clinit>(Nd4jCpu.java:62)
        at java.lang.Class.forName0(Native Method)
        at java.lang.Class.forName(Class.java:264)
        at org.nd4j.nativeblas.NativeOpsHolder.<init>(NativeOpsHolder.java:29)
        ... 19 more
Caused by: java.lang.UnsatisfiedLinkError: Native Library /home/yarn/.javacpp/cache/blob_a87e49f9a475a9dc4296f6afbc3ae171dc821d19/org/nd4j/nativeblas/linux-x86_64/libjnind4jcpu.so already loaded in another classloader
        at java.lang.ClassLoader.loadLibrary0(ClassLoader.java:1903)
        at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1822)
        at java.lang.Runtime.load0(Runtime.java:809)
        at java.lang.System.load(System.java:1086)
        at org.bytedeco.javacpp.Loader.loadLibrary(Loader.java:943)
 
I’m no expert in class loading at all but have tried a number of ways to overcome this based on the docs and dynamic classloading
 
1.       Putting the jars that contain the native libraries in the flink/lib directory so they are always loaded on JobManager startup and removing them from the deployed flink job
2.       Some code in the open / close methods of my mapper to try and dispose of the native library
NativeLibrary lib = NativeLibrary.getInstance(Loader.getCacheDir() + "/cache/nd4j-native-0.8.0-" + Loader.getPlatform() +  ".jar/org/nd4j/nativeblas/" + Loader.getPlatform() + "/libjnind4jcpu.so");
this.model = null;

if (null != lib) {
    System.out.println("found lib - removing");
    lib.dispose();
    System.gc();
}
This sort of works in finding the library (on my mac running locally) but as from the stacktrace above, on deployment to yarn some random value is placed after the /cache dir which I don’t know how to get a handle on to be able to construc the correct library location (and just using the library name as the jni docs suggest for NativeLibrary.getInstance fails to find the library)
 
Neither of the above approaches work, so when my job fails I can’t cancel and resubmit as it just fails with the same stack trace. The only way I can get it to run again is the cancel all other jobs running on JobManager, killing the JM yarn session, creating a new Yarn session JM, resubmitting flink job – which is a real pain. Ideally I would like to stop the exception in the first place, but as I can’t figure out how to get logging appearing in my yarn logs either (for debug) I’m at a bit of a loss!
 
Any pointers, suggestions please??
 
Many thanks
Conrad
 


SecureData, combating cyber threats


The information contained in this message or any of its attachments may be privileged and confidential and intended for the exclusive use of the intended recipient. If you are not the intended recipient any disclosure, reproduction, distribution or other dissemination or use of this communications is strictly prohibited. The views expressed in this email are those of the individual and not necessarily of SecureData Europe Ltd. Any prices quoted are only valid if followed up by a formal written quote.

SecureData Europe Limited. Registered in England & Wales 04365896. Registered Address: SecureData House, Hermitage Court, Hermitage Lane, Maidstone, Kent, ME16 9NT


Loading...