Hi everyone,
I posted the following message on the Ignite Users list, but
stephendarlington
<http://apache-ignite-users.70518.x6.nabble.com/template/NamlServlet.jtp?macro=user_nodes&user=1881>
suggested that it would be better to post it in the dev list.
Original message:
After the new release (2.7.0), I have been playing around with the machine
learning algorithms a bit.
We have some data in a cache created with the "withKeepBinary()" option, and
I wanted
to test if the machine learning algos would work with such a cache. I tried,
but it fails with the following stacktrace:
org.apache.ignite.IgniteException: testType
at
org.apache.ignite.internal.processors.closure.GridClosureProcessor$C2.execute(GridClosureProcessor.java:1858)
at
org.apache.ignite.internal.processors.job.GridJobWorker$2.call(GridJobWorker.java:568)
at
org.apache.ignite.internal.util.IgniteUtils.wrapThreadLoader(IgniteUtils.java:6816)
at
org.apache.ignite.internal.processors.job.GridJobWorker.execute0(GridJobWorker.java:562)
at
org.apache.ignite.internal.processors.job.GridJobWorker.body(GridJobWorker.java:491)
at
org.apache.ignite.internal.util.worker.GridWorker.run(GridWorker.java:120)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.ignite.binary.BinaryInvalidTypeException: testType
at
org.apache.ignite.internal.binary.BinaryContext.descriptorForTypeId(BinaryContext.java:707)
at
org.apache.ignite.internal.binary.BinaryReaderExImpl.deserialize0(BinaryReaderExImpl.java:1757)
at
org.apache.ignite.internal.binary.BinaryReaderExImpl.deserialize(BinaryReaderExImpl.java:1716)
at
org.apache.ignite.internal.binary.BinaryObjectImpl.deserializeValue(BinaryObjectImpl.java:798)
at
org.apache.ignite.internal.binary.BinaryObjectImpl.value(BinaryObjectImpl.java:143)
at
org.apache.ignite.internal.processors.cache.CacheObjectUtils.unwrapBinary(CacheObjectUtils.java:177)
at
org.apache.ignite.internal.processors.cache.CacheObjectUtils.unwrapBinaryIfNeeded(CacheObjectUtils.java:39)
at
org.apache.ignite.internal.processors.cache.query.GridCacheQueryManager$ScanQueryIterator.advance(GridCacheQueryManager.java:3063)
at
org.apache.ignite.internal.processors.cache.query.GridCacheQueryManager$ScanQueryIterator.onHasNext(GridCacheQueryManager.java:2965)
at
org.apache.ignite.internal.util.GridCloseableIteratorAdapter.hasNextX(GridCloseableIteratorAdapter.java:53)
at
org.apache.ignite.internal.util.lang.GridIteratorAdapter.hasNext(GridIteratorAdapter.java:45)
at
org.apache.ignite.ml.dataset.impl.cache.util.ComputeUtils.computeCount(ComputeUtils.java:313)
at
org.apache.ignite.ml.dataset.impl.cache.util.ComputeUtils.computeCount(ComputeUtils.java:300)
at
org.apache.ignite.ml.dataset.impl.cache.util.ComputeUtils.lambda$initContext$9b68d858$1(ComputeUtils.java:222)
at
org.apache.ignite.ml.dataset.impl.cache.util.ComputeUtils.lambda$affinityCallWithRetries$b46c4136$1(ComputeUtils.java:90)
at
org.apache.ignite.internal.processors.closure.GridClosureProcessor$C2.execute(GridClosureProcessor.java:1855)
... 8 common frames omitted
Caused by: java.lang.ClassNotFoundException: testType
at java.net.URLClassLoader.findClass(URLClassLoader.java:381)
at java.lang.ClassLoader.loadClass(ClassLoader.java:424)
at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:338)
at java.lang.ClassLoader.loadClass(ClassLoader.java:357)
at java.lang.Class.forName0(Native Method)
at java.lang.Class.forName(Class.java:348)
at
org.apache.ignite.internal.util.IgniteUtils.forName(IgniteUtils.java:8771)
at
org.apache.ignite.internal.MarshallerContextImpl.getClass(MarshallerContextImpl.java:349)
at
org.apache.ignite.internal.binary.BinaryContext.descriptorForTypeId(BinaryContext.java:698)
... 23 common frames omitted
Debugging, I found the source of the error, at some point you are just
taking the
name of the upstreamCache (where the data resides), and creating a new
IgniteCache
object using such name before copying the data to a dataset cache. However,
you
are not using the keepBinary property of the original cache. I hardcoded the
"withKeepBinary()" to the following lines:
https://github.com/apache/ignite/blob/2.7.0/modules/ml/src/main/java/org/apache/ignite/ml/dataset/impl/cache/util/ComputeUtils.java#L162
https://github.com/apache/ignite/blob/2.7.0/modules/ml/src/main/java/org/apache/ignite/ml/dataset/impl/cache/util/ComputeUtils.java#L215
https://github.com/apache/ignite/blob/2.7.0/modules/ml/src/main/java/org/apache/ignite/ml/dataset/impl/cache/CacheBasedDatasetBuilder.java#L99
The previous made it work. I tried to retrieve the keep binary property from
the
upstreamCache, but I was not able to find the right method to obtain it (I
saw the property is
stored in the operation context field (opCtx), but it is private and cannot
be
accessed from the lines I modified)
My example code is available at:
https://gist.github.com/otorreno/ca6c5347c1bbde2d4fedd02b51d02cbb
Any plans on making the machine learning algorithms work with caches with
keepBinary set to true?
--
Sent from: http://apache-ignite-developers.2346864.n4.nabble.com/
