You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@ignite.apache.org by otorreno <os...@shapelets.io> on 2019/01/02 07:20:16 UTC
Ignite ML withKeepBinary cache
Hi everyone,
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-users.70518.x6.nabble.com/
Re: Ignite ML withKeepBinary cache
Posted by Yuriy Babak <y....@gmail.com>.
Hi all,
Ticket 10700 [1] is resolved, this ticked added support of training models
over a cache with binary objects(cache with enabled keepBinary flag) for
more details please take a look the mentioned ticked or added example [2].
[1] - https://issues.apache.org/jira/browse/IGNITE-10700
[2] - org.apache.ignite.examples.ml.TrainingWithBinaryObjectExample
Best regards,
Yuriy Babak
пн, 21 янв. 2019 г. в 11:13, otorreno <os...@shapelets.io>:
> Thanks Ilya,
>
> I got the link to the JIRA ticket in the Ignite Devs mailing list. In fact
> I
> already included a comment in the ticket.
>
> I got a response from Alexey Zinoviev in the Dev list too, and I am now
> waiting to receive a further update from him on this matter.
>
> Regards,
>
>
>
> --
> Sent from: http://apache-ignite-users.70518.x6.nabble.com/
>
Re: Ignite ML withKeepBinary cache
Posted by otorreno <os...@shapelets.io>.
Thanks Ilya,
I got the link to the JIRA ticket in the Ignite Devs mailing list. In fact I
already included a comment in the ticket.
I got a response from Alexey Zinoviev in the Dev list too, and I am now
waiting to receive a further update from him on this matter.
Regards,
--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/
Re: Ignite ML withKeepBinary cache
Posted by "ilya.kasnacheev" <il...@gmail.com>.
Hello!
I can see there is a ticket in JIRA:
https://issues.apache.org/jira/browse/IGNITE-10700
If anything is not there please update it.
Regards,
--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/
Re: Ignite ML withKeepBinary cache
Posted by Stephen Darlington <st...@gridgain.com>.
That’s a great investigation! I think the developer mailing list (http://apache-ignite-developers.2346864.n4.nabble.com) would be a better place to discuss the best way to fix it, though.
Regards,
Stephen
> On 2 Jan 2019, at 07:20, otorreno <os...@shapelets.io> wrote:
>
> Hi everyone,
>
> 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-users.70518.x6.nabble.com/