[jira] [Created] (IGNITE-12161) [ML] Add support of different classloaders for client code
Alexey Platonov created IGNITE-12161: Summary: [ML] Add support of different classloaders for client code Key: IGNITE-12161 URL: https://issues.apache.org/jira/browse/IGNITE-12161 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Currently we support just one classloader for unknown classes on server side. Potentially client can handle several classloaders and we should support this situation. -- This message was sent by Atlassian Jira (v8.3.2#803003)
[jira] [Created] (IGNITE-12157) [ML] Implement distributed ROC AUC computation
Alexey Platonov created IGNITE-12157: Summary: [ML] Implement distributed ROC AUC computation Key: IGNITE-12157 URL: https://issues.apache.org/jira/browse/IGNITE-12157 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Currently, we don't have valid ROC AUC computation at all. We should implement it with predict proba interface in models. It is desirable that ROC AUC computation will be distributed. -- This message was sent by Atlassian Jira (v8.3.2#803003)
[jira] [Created] (IGNITE-12156) [ML] Add meta information to models
Alexey Platonov created IGNITE-12156: Summary: [ML] Add meta information to models Key: IGNITE-12156 URL: https://issues.apache.org/jira/browse/IGNITE-12156 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Current models don't contain any information about their learning process or a features meta-information or information about the type of model (for example RegressionTree doesn't differ from ClassificationTree in the interface). It leads to extra work in other APIs. For example, we cannot define automatically set of metrics for a user-passed model in Evaluator. -- This message was sent by Atlassian Jira (v8.3.2#803003)
[jira] [Created] (IGNITE-12155) [ML] Implementation of distributed estimator
Alexey Platonov created IGNITE-12155: Summary: [ML] Implementation of distributed estimator Key: IGNITE-12155 URL: https://issues.apache.org/jira/browse/IGNITE-12155 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Currently, we don't have an ability to compute metrics in a distributed manner and it leads to reading all data from partitions to a client using Qursor. We should develop a framework for distributed computing of metrics. -- This message was sent by Atlassian Jira (v8.3.2#803003)
[jira] [Created] (IGNITE-12024) [ML] PeerClassloading for ml related lambdas
Alexey Platonov created IGNITE-12024: Summary: [ML] PeerClassloading for ml related lambdas Key: IGNITE-12024 URL: https://issues.apache.org/jira/browse/IGNITE-12024 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov The current implementation of peerClassloading doesn't fulfill all ml requirements. We want to provide ability of implementation custom vectorizers and preprocessors. -- This message was sent by Atlassian JIRA (v7.6.14#76016)
[jira] [Created] (IGNITE-11675) [ML] Create additional examples for linear regressions, knn and kmeans
Alexey Platonov created IGNITE-11675: Summary: [ML] Create additional examples for linear regressions, knn and kmeans Key: IGNITE-11675 URL: https://issues.apache.org/jira/browse/IGNITE-11675 Project: Ignite Issue Type: Improvement Components: documentation, ml Reporter: Alexey Platonov Assignee: Alexey Platonov We should create additional examples for linear regressions, knn and kmeans in examples module for ebook about ML in Ignite. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11664) [ML] Use Double.NaN as default values for missing values in Vector
Alexey Platonov created IGNITE-11664: Summary: [ML] Use Double.NaN as default values for missing values in Vector Key: IGNITE-11664 URL: https://issues.apache.org/jira/browse/IGNITE-11664 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Currently, we use 0.0 value for default values in vectors if a value is missing. But this way contradicts to preprocessors politics where for missing values Double.NaN is using. Moreover, Double.NaN is a more convenient value for missing feature values. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11647) [ML] ML Vectors should work with all Serializable objects besides double
Alexey Platonov created IGNITE-11647: Summary: [ML] ML Vectors should work with all Serializable objects besides double Key: IGNITE-11647 URL: https://issues.apache.org/jira/browse/IGNITE-11647 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Such refactoring allows us to reduce hierarchy parallelism between preprocessors and dataset trainers. Also, support of types besides double allows improving some algorithms like decision trees. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11642) [ML] Umbrella: API for Feature/Label extracting (part 2)
Alexey Platonov created IGNITE-11642: Summary: [ML] Umbrella: API for Feature/Label extracting (part 2) Key: IGNITE-11642 URL: https://issues.apache.org/jira/browse/IGNITE-11642 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Fix For: 2.8 Replace current lambdas with fixed API -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11582) [ML] Pipelines should work with Vectorizers
Alexey Platonov created IGNITE-11582: Summary: [ML] Pipelines should work with Vectorizers Key: IGNITE-11582 URL: https://issues.apache.org/jira/browse/IGNITE-11582 Project: Ignite Issue Type: Bug Components: ml Reporter: Alexey Platonov Currently Pipelines are implemented using feature/label extraction functions with generic label value. We should adapt pipelines for vectorizers (maybe after this ticket - IGNITE-11481). -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11580) [ML] Evaluators should accept Vectorizers
Alexey Platonov created IGNITE-11580: Summary: [ML] Evaluators should accept Vectorizers Key: IGNITE-11580 URL: https://issues.apache.org/jira/browse/IGNITE-11580 Project: Ignite Issue Type: Bug Components: ml Reporter: Alexey Platonov Currently evaluation API uses old interface with separated feature-label extractors. In context of IGNITE-11449 we should change this API. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11581) [ML] Adapt tutorial to new vectorizer API
Alexey Platonov created IGNITE-11581: Summary: [ML] Adapt tutorial to new vectorizer API Key: IGNITE-11581 URL: https://issues.apache.org/jira/browse/IGNITE-11581 Project: Ignite Issue Type: Bug Components: ml Reporter: Alexey Platonov Currently tutorial uses old feature-labels extraction API -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11561) [ML] IgniteDistributedModel for XGBoost doesn't work in example
Alexey Platonov created IGNITE-11561: Summary: [ML] IgniteDistributedModel for XGBoost doesn't work in example Key: IGNITE-11561 URL: https://issues.apache.org/jira/browse/IGNITE-11561 Project: Ignite Issue Type: Bug Components: ml Affects Versions: 2.7, 2.8 Reporter: Alexey Platonov Assignee: Anton Dmitriev Distributed inference model for XGBoost doesn't work in example (XGBoostModelParserExample). It always returns same value. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11504) [ML] Preprocessor trainers should support new feature-label extraction API
Alexey Platonov created IGNITE-11504: Summary: [ML] Preprocessor trainers should support new feature-label extraction API Key: IGNITE-11504 URL: https://issues.apache.org/jira/browse/IGNITE-11504 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Problem is same as feature extractors serialization bug. We should narrow our API. (see parent task) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11481) [ML] Prototype of DatasetRow for Vectorizer
Alexey Platonov created IGNITE-11481: Summary: [ML] Prototype of DatasetRow for Vectorizer Key: IGNITE-11481 URL: https://issues.apache.org/jira/browse/IGNITE-11481 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Vectorizer shold produce DatasetRow object that can contains columns with different types (double, string, etc.). It needs for preprocessors working with non-double values. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11480) [ML] Use only Vectorizer API in DatasetTrainer API
Alexey Platonov created IGNITE-11480: Summary: [ML] Use only Vectorizer API in DatasetTrainer API Key: IGNITE-11480 URL: https://issues.apache.org/jira/browse/IGNITE-11480 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Use only Vectorizer API in DatasetTrainer API to avoid problems with user classes serialization. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11479) [ML] Use new vectorizer API in PartitionDatasetBuilders
Alexey Platonov created IGNITE-11479: Summary: [ML] Use new vectorizer API in PartitionDatasetBuilders Key: IGNITE-11479 URL: https://issues.apache.org/jira/browse/IGNITE-11479 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov We need to exclude current feature extractors from partition building API and replace old extractors with new vectorizer. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11478) [ML] Use new vectorizer API in Trainers
Alexey Platonov created IGNITE-11478: Summary: [ML] Use new vectorizer API in Trainers Key: IGNITE-11478 URL: https://issues.apache.org/jira/browse/IGNITE-11478 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov We should rewrite current trainers - exclude all "free"-feature/labels extractors from APIs and use new vectorizer in them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11476) [ML] Use new feature extraction API in examples
Alexey Platonov created IGNITE-11476: Summary: [ML] Use new feature extraction API in examples Key: IGNITE-11476 URL: https://issues.apache.org/jira/browse/IGNITE-11476 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Introduce new feature/label extraction API to all examples. These examples should work on binary builds without sharing additional jars to libs directory (except ml-jar). -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11477) [ML] Create tests for ML algorithms stability check against binary builds
Alexey Platonov created IGNITE-11477: Summary: [ML] Create tests for ML algorithms stability check against binary builds Key: IGNITE-11477 URL: https://issues.apache.org/jira/browse/IGNITE-11477 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov After new feature API creation we should create tests for ML algorithms stability check against binary builds (or on other JVMs without common classpath). All new algorithms should be delivered with such test. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11475) [ML] Vectorizer API prototype with POC
Alexey Platonov created IGNITE-11475: Summary: [ML] Vectorizer API prototype with POC Key: IGNITE-11475 URL: https://issues.apache.org/jira/browse/IGNITE-11475 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov We need to create a prototype of API for features/labels extraction and introduce it to one or two already existing examples. This prototype should show that new API works on binary builds. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11401) [ML] Labmdas doesn't work in binary builds
Alexey Platonov created IGNITE-11401: Summary: [ML] Labmdas doesn't work in binary builds Key: IGNITE-11401 URL: https://issues.apache.org/jira/browse/IGNITE-11401 Project: Ignite Issue Type: Bug Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Fix For: 2.8 Current lambdas for feature extraction doesn't work in binary builds and fail with such errors: Caused by: class org.apache.ignite.binary.BinaryObjectException: Failed to deserialize object [typeName=org.apache.ignite.ml.regressions.linear.FeatureExtractorWrapper] at org.apache.ignite.internal.binary.BinaryClassDescriptor.read(BinaryClassDescriptor.java:913) at org.apache.ignite.internal.binary.BinaryReaderExImpl.deserialize0(BinaryReaderExImpl.java:1763) at org.apache.ignite.internal.binary.BinaryReaderExImpl.deserialize(BinaryReaderExImpl.java:1715) at org.apache.ignite.internal.binary.BinaryReaderExImpl.readField(BinaryReaderExImpl.java:1983) at org.apache.ignite.internal.binary.BinaryFieldAccessor$DefaultFinalClassAccessor.read0(BinaryFieldAccessor.java:703) at org.apache.ignite.internal.binary.BinaryFieldAccessor.read(BinaryFieldAccessor.java:188) ... 40 more Caused by: class org.apache.ignite.binary.BinaryObjectException: Failed to read field [name=featureExtractor] at org.apache.ignite.internal.binary.BinaryFieldAccessor.read(BinaryFieldAccessor.java:192) at org.apache.ignite.internal.binary.BinaryClassDescriptor.read(BinaryClassDescriptor.java:874) ... 45 more Caused by: class org.apache.ignite.binary.BinaryObjectException: Failed to deserialize object [typeName=java.lang.invoke.SerializedLambda] at org.apache.ignite.internal.binary.BinaryClassDescriptor.read(BinaryClassDescriptor.java:913) at org.apache.ignite.internal.binary.BinaryReaderExImpl.deserialize0(BinaryReaderExImpl.java:1763) at org.apache.ignite.internal.binary.BinaryReaderExImpl.deserialize(BinaryReaderExImpl.java:1715) at org.apache.ignite.internal.binary.BinaryReaderExImpl.readField(BinaryReaderExImpl.java:1983) at org.apache.ignite.internal.binary.BinaryFieldAccessor$DefaultFinalClassAccessor.read0(BinaryFieldAccessor.java:703) at org.apache.ignite.internal.binary.BinaryFieldAccessor.read(BinaryFieldAccessor.java:188) ... 46 more Caused by: class org.apache.ignite.binary.BinaryObjectException: Failed to read field [name=capturedArgs] at org.apache.ignite.internal.binary.BinaryFieldAccessor.read(BinaryFieldAccessor.java:192) at org.apache.ignite.internal.binary.BinaryClassDescriptor.read(BinaryClassDescriptor.java:874) ... 51 more Caused by: class org.apache.ignite.binary.BinaryInvalidTypeException: org.apache.ignite.examples.ml.regression.linear.StaticFun -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11372) [ML] Fix javadoc in VectorWithDistributionId
Alexey Platonov created IGNITE-11372: Summary: [ML] Fix javadoc in VectorWithDistributionId Key: IGNITE-11372 URL: https://issues.apache.org/jira/browse/IGNITE-11372 Project: Ignite Issue Type: Bug Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov "mvn initialize -Pjavadoc" fails with error "[ERROR] Failed to execute goal org.apache.maven.plugins:maven-antrun-plugin:1.7:run (javadoc-postprocessing-new) on project apache-ignite: An Ant BuildException has occured: Execution failed due to: Class doesn't have description in file: <..>/ml/util/generators/primitives/vector/VectorGeneratorsFamily.VectorWithDistributionId.html" -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11241) [ML] blas NoClassDefFoundError exception
Alexey Platonov created IGNITE-11241: Summary: [ML] blas NoClassDefFoundError exception Key: IGNITE-11241 URL: https://issues.apache.org/jira/browse/IGNITE-11241 Project: Ignite Issue Type: Bug Components: ml Affects Versions: 2.7, 2.6, 2.5, 2.4, 2.3, 2.2, 2.1, 2.0, 2.8 Reporter: Alexey Platonov Assignee: Alexey Platonov Fix For: 2.8 Caused by: java.lang.NoClassDefFoundError: org/netlib/blas/Dnrm2 at com.github.fommil.netlib.F2jBLAS.dnrm2(F2jBLAS.java:121) at org.apache.ignite.ml.math.isolve.lsqr.LSQROnHeap.lambda$bnorm$7b55d622$1(LSQROnHeap.java:53) at org.apache.ignite.ml.dataset.Dataset.lambda$computeWithCtx$96d91738$1(Dataset.java:143) at org.apache.ignite.ml.dataset.impl.cache.CacheBasedDataset.lambda$computeWithCtx$62946a87$1(CacheBasedDataset.java:112) at org.apache.ignite.ml.dataset.impl.cache.util.ComputeUtils.lambda$affinityCallWithRetries$b46c4136$1(ComputeUtils.java:90) at org.apache.ignite.internal.processors.closure.GridClosureProcessor$C2.execute(GridClosureProcessor.java:1855) 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) ... 11 more -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11234) [ML] Add fillCache method to DataStreamGenerator
Alexey Platonov created IGNITE-11234: Summary: [ML] Add fillCache method to DataStreamGenerator Key: IGNITE-11234 URL: https://issues.apache.org/jira/browse/IGNITE-11234 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov We need to add fillCache(sampleSize, cache) method to DataStreamGenerator for a more convenient interface in terms of cache filling + add example for such method. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11232) [ML] Random Forest, NodeId exception (Failed to serialize object)
Alexey Platonov created IGNITE-11232: Summary: [ML] Random Forest, NodeId exception (Failed to serialize object) Key: IGNITE-11232 URL: https://issues.apache.org/jira/browse/IGNITE-11232 Project: Ignite Issue Type: Bug Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Caused by: class org.apache.ignite.IgniteCheckedException: Failed to serialize object: IgniteBiTuple [val1=0, val2=1] at org.apache.ignite.internal.marshaller.optimized.OptimizedMarshaller.marshal0(OptimizedMarshaller.java:207) at org.apache.ignite.marshaller.AbstractNodeNameAwareMarshaller.marshal(AbstractNodeNameAwareMarshaller.java:58) at org.apache.ignite.internal.util.IgniteUtils.marshal(IgniteUtils.java:10222) at org.apache.ignite.internal.binary.BinaryWriterExImpl.marshal0(BinaryWriterExImpl.java:196) ... 99 more Caused by: java.io.IOException: Externalizable class doesn't have default constructor: class org.apache.ignite.ml.tree.randomforest.data.NodeId at org.apache.ignite.internal.marshaller.optimized.OptimizedClassDescriptor.(OptimizedClassDescriptor.java:408) at org.apache.ignite.internal.marshaller.optimized.OptimizedMarshallerUtils.classDescriptor(OptimizedMarshallerUtils.java:210) at org.apache.ignite.internal.marshaller.optimized.OptimizedObjectOutputStream.writeObject0(OptimizedObjectOutputStream.java:201) at org.apache.ignite.internal.marshaller.optimized.OptimizedObjectOutputStream.writeObjectOverride(OptimizedObjectOutputStream.java:159) at java.io.ObjectOutputStream.writeObject(ObjectOutputStream.java:344) at org.apache.ignite.internal.marshaller.optimized.OptimizedMarshaller.marshal0(OptimizedMarshaller.java:202) ... 102 more Caused by: java.lang.NoSuchMethodException: org.apache.ignite.ml.tree.randomforest.data.NodeId.() at java.lang.Class.getConstructor0(Class.java:3082) at java.lang.Class.getDeclaredConstructor(Class.java:2178) at org.apache.ignite.internal.marshaller.optimized.OptimizedClassDescriptor.(OptimizedClassDescriptor.java:403) ... 107 more -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11192) [ML] Use nd4j for matrix inversions and determinants
Alexey Platonov created IGNITE-11192: Summary: [ML] Use nd4j for matrix inversions and determinants Key: IGNITE-11192 URL: https://issues.apache.org/jira/browse/IGNITE-11192 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov >From optimization point of view we should use matrix inversions and >determinant computations of dl4j instead of own realization. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-11133) Refactoring of compatibility framework
Alexey Platonov created IGNITE-11133: Summary: Refactoring of compatibility framework Key: IGNITE-11133 URL: https://issues.apache.org/jira/browse/IGNITE-11133 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov We need to refactor current compatibility framework with the aim of reducing maven management logic, simplification adding new versions to check. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10793) [ML] Create comprehensive example for dataset generators
Alexey Platonov created IGNITE-10793: Summary: [ML] Create comprehensive example for dataset generators Key: IGNITE-10793 URL: https://issues.apache.org/jira/browse/IGNITE-10793 Project: Ignite Issue Type: Sub-task Components: ml Reporter: Alexey Platonov Assignee: Alexey Platonov Fix For: 2.8 We need to create well documented example for generators. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10727) [ML] InfModel and Model merging
Alexey Platonov created IGNITE-10727: Summary: [ML] InfModel and Model merging Key: IGNITE-10727 URL: https://issues.apache.org/jira/browse/IGNITE-10727 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov Assignee: Anton Dmitriev Fix For: 2.8 Currently "InfModel" and "Model" provide parallel architecture in terms of using of "InfModels" in after-learning steps like compositions, estimations etc. I propose to move "InfModel" to top of models hierarchy and rename it to just "Model" and current "Model" rename to "LearnedModel/LocalModel/Your_ad_may_be_here". So in this way "InfModel" will be just model with apply-function for generic I-O values and "Model" will be serializable thing on Vector-Double. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10549) [ML] LogisticRegressionSGDTrainerTest fails with BLAS
Alexey Platonov created IGNITE-10549: Summary: [ML] LogisticRegressionSGDTrainerTest fails with BLAS Key: IGNITE-10549 URL: https://issues.apache.org/jira/browse/IGNITE-10549 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov Assignee: Alexey Platonov LogisticRegressionSGDTrainerTest fails with BLAS on case with 3 partitions. Investigation is needed. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10517) [ML] Merge inference and learning architectures
Alexey Platonov created IGNITE-10517: Summary: [ML] Merge inference and learning architectures Key: IGNITE-10517 URL: https://issues.apache.org/jira/browse/IGNITE-10517 Project: Ignite Issue Type: Sub-task Components: ml Reporter: Alexey Platonov Fix For: 2.8 We need to review of inference framework and try to merge it with core part on ML library in terms of reusing imported models in all ML steps excluding learning (model estimation, compositions etc.) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10503) Meta information for vectors
Alexey Platonov created IGNITE-10503: Summary: Meta information for vectors Key: IGNITE-10503 URL: https://issues.apache.org/jira/browse/IGNITE-10503 Project: Ignite Issue Type: Improvement Components: ml Reporter: Alexey Platonov We need to design and implement vector meta-information like feature names, bagging information, etc -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10440) Analyse test suites for possible acceleration
Alexey Platonov created IGNITE-10440: Summary: Analyse test suites for possible acceleration Key: IGNITE-10440 URL: https://issues.apache.org/jira/browse/IGNITE-10440 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10195) Cannot create caches with different names but with same indexed types and schema name
Alexey Platonov created IGNITE-10195:
Summary: Cannot create caches with different names but with same
indexed types and schema name
Key: IGNITE-10195
URL: https://issues.apache.org/jira/browse/IGNITE-10195
Project: Ignite
Issue Type: Improvement
Reporter: Alexey Platonov
Cannot create caches with different names but with same indexed types and
schema name. For example, such code will throw exception
"javax.cache.CacheException: Table already exists: VALUE".
{color:#9876aa}node{color}.createCache({color:#cc7832}new
{color}CacheConfiguration()
.setName({color:#9876aa}"PERSON_1"{color})
.setIndexedTypes(Key.{color:#cc7832}class,{color}
Person.{color:#cc7832}class{color})
.setSqlSchema(QueryUtils.{color:#9876aa}DFLT_SCHEMA{color})){color:#cc7832};
{color}{color:#cc7832}
{color}{color:#9876aa}node{color}.createCache({color:#cc7832}new
{color}CacheConfiguration()
.setName({color:#9876aa}"PERSON_2"{color})
.setIndexedTypes({color:#cc7832}Key.class, Person.class{color})
.setSqlSchema(QueryUtils.{color:#9876aa}DFLT_SCHEMA{color})){color:#cc7832};{color}
If I set table name manually by setQueryEntities(...) then
"javax.cache.CacheException: Index already exists: PERSON_ORGID_IDX" wil be
thrown (Value has field with "origId" and annotation
{color:#bbb529}@QuerySqlField{color}({color:#d0d0ff}index {color}=
{color:#cc7832}true{color})). Here is definition of Person class:
{color:#cc7832}public static class {color}PersonKey {
{color:#bbb529}@QuerySqlField
{color} {color:#cc7832}public long
{color}{color:#9876aa}id{color}{color:#cc7832};
{color}{color:#cc7832}
{color} {color:#629755}/**
{color}{color:#629755} * Constructor.
{color}{color:#629755} *
{color}{color:#629755} * {color}{color:#629755}@param {color}{color:#8a653b}id
{color}{color:#629755}ID.
{color}{color:#629755} */
{color} {color:#ffc66d}PersonKey{color}({color:#cc7832}long {color}id) {
{color:#cc7832}this{color}.{color:#9876aa}id {color}= id{color:#cc7832};
{color} }
{color:#629755}/** {{color}{color:#629755}@inheritDoc{color}{color:#629755}} */
{color} {color:#bbb529}@Override {color}{color:#cc7832}public int
{color}{color:#ffc66d}hashCode{color}() {
{color:#cc7832}return
{color}({color:#cc7832}int{color}){color:#9876aa}id{color}{color:#cc7832};
{color} }
{color:#629755}/** {{color}{color:#629755}@inheritDoc{color}{color:#629755}} */
{color} {color:#bbb529}@Override {color}{color:#cc7832}public boolean
{color}{color:#ffc66d}equals{color}(Object obj) {
{color:#cc7832}return {color}obj != {color:#cc7832}null {color}&& obj
{color:#cc7832}instanceof {color}PersonKey &&
(F.eq({color:#9876aa}id{color}{color:#cc7832},
{color}((PersonKey)obj).{color:#9876aa}id{color})){color:#cc7832};
{color} }
}
Such behavior seems to be usability bug. Why I cannot create two caches with
different names but with same indexed values.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
[jira] [Created] (IGNITE-10194) ZookeeperDiscoverySpiTest fails on testLocalAuthenticationFails
Alexey Platonov created IGNITE-10194: Summary: ZookeeperDiscoverySpiTest fails on testLocalAuthenticationFails Key: IGNITE-10194 URL: https://issues.apache.org/jira/browse/IGNITE-10194 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov We need to unmute test (remove "_" prefix) and fix this test -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10141) Cache 3 tests optimization
Alexey Platonov created IGNITE-10141: Summary: Cache 3 tests optimization Key: IGNITE-10141 URL: https://issues.apache.org/jira/browse/IGNITE-10141 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov We need to investigate how to optimize these tests: GridCacheInterceptorTransactionalRebalanceTest.testRebalanceUpdate GridCacheInterceptorTransactionalRebalanceTest.testRebalanceRemoveInvoke GridCacheInterceptorTransactionalRebalanceTest.testPutIfAbsent GridCacheInterceptorTransactionalRebalanceTest.testRebalanceUpdateInvoke IgniteCacheGroupsTest.testRestartsAndCacheCreateDestroy GridCacheInterceptorTransactionalRebalanceTest.testRebalanceRemove GridCacheInterceptorTransactionalRebalanceTest.testGetAndPut IgniteCacheGroupsTest.testStartManyCaches GridCacheVersionTopologyChangeTest.testVersionIncreaseTx GridCacheVersionTopologyChangeTest.testVersionIncreaseAtomic GridCacheValueConsistencyTransactionalSelfTest.testPutRemoveConsistencyMultithreaded GridCacheInterceptorAtomicRebalanceTest.testRebalanceUpdateInvoke GridCacheInterceptorAtomicRebalanceTest.testGetAndPut and optimize them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10106) Cache 5 test suite optimization
Alexey Platonov created IGNITE-10106: Summary: Cache 5 test suite optimization Key: IGNITE-10106 URL: https://issues.apache.org/jira/browse/IGNITE-10106 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov We need to investigate how to optimize these tests: |CacheSerializableTransactionsTest.testGetRemoveTxNearCache2| |[CacheSerializableTransactionsTest.testGetRemoveTxNearCache1|https://ci.ignite.apache.org/viewLog.html?currentGroup=test&scope=org.apache.ignite.testsuites.IgniteCacheTestSuite5%23teamcity%23org.apache.ignite.internal.processors.cache%23teamcity%23CacheSerializableTransactionsTest&pager.currentPage=1&order=DURATION_DESC&recordsPerPage=20&filterText=&status=&buildTypeId=IgniteTests24Java8_Cache5&buildId=2160072&tab=testsInfo]| |CacheSerializableTransactionsTest.testRandomOperations| |[CacheSerializableTransactionsTest.testIncrementTxRestart|https://ci.ignite.apache.org/viewLog.html?currentGroup=test&scope=org.apache.ignite.testsuites.IgniteCacheTestSuite5%23teamcity%23org.apache.ignite.internal.processors.cache%23teamcity%23CacheSerializableTransactionsTest&pager.currentPage=1&order=DURATION_DESC&recordsPerPage=20&filterText=&status=&buildTypeId=IgniteTests24Java8_Cache5&buildId=2160072&tab=testsInfo]| |CacheSerializableTransactionsTest.testConcurrentUpdateNoDeadlockNodeRestart| |[CacheSerializableTransactionsTest.testConcurrentUpdateNoDeadlockFromClientsNodeRestart|https://ci.ignite.apache.org/viewLog.html?currentGroup=test&scope=org.apache.ignite.testsuites.IgniteCacheTestSuite5%23teamcity%23org.apache.ignite.internal.processors.cache%23teamcity%23CacheSerializableTransactionsTest&pager.currentPage=1&order=DURATION_DESC&recordsPerPage=20&filterText=&status=&buildTypeId=IgniteTests24Java8_Cache5&buildId=2160072&tab=testsInfo]| |[CacheSerializableTransactionsTest.testConcurrentUpdateNoDeadlockWithNonSerializable|https://ci.ignite.apache.org/viewLog.html?currentGroup=test&scope=org.apache.ignite.testsuites.IgniteCacheTestSuite5%23teamcity%23org.apache.ignite.internal.processors.cache%23teamcity%23CacheSerializableTransactionsTest&pager.currentPage=1&order=DURATION_DESC&recordsPerPage=20&filterText=&status=&buildTypeId=IgniteTests24Java8_Cache5&buildId=2160072&tab=testsInfo]| |[CacheSerializableTransactionsTest.testConcurrentUpdateNoDeadlockGetPut|https://ci.ignite.apache.org/viewLog.html?currentGroup=test&scope=org.apache.ignite.testsuites.IgniteCacheTestSuite5%23teamcity%23org.apache.ignite.internal.processors.cache%23teamcity%23CacheSerializableTransactionsTest&pager.currentPage=1&order=DURATION_DESC&recordsPerPage=20&filterText=&status=&buildTypeId=IgniteTests24Java8_Cache5&buildId=2160072&tab=testsInfo]| |[CacheSerializableTransactionsTest.testConcurrentUpdateNoDeadlockFromClients|https://ci.ignite.apache.org/viewLog.html?currentGroup=test&scope=org.apache.ignite.testsuites.IgniteCacheTestSuite5%23teamcity%23org.apache.ignite.internal.processors.cache%23teamcity%23CacheSerializableTransactionsTest&pager.currentPage=1&order=DURATION_DESC&recordsPerPage=20&filterText=&status=&buildTypeId=IgniteTests24Java8_Cache5&buildId=2160072&tab=testsInfo]| |[CacheSerializableTransactionsTest.testConcurrentUpdateNoDeadlock|https://ci.ignite.apache.org/viewLog.html?currentGroup=test&scope=org.apache.ignite.testsuites.IgniteCacheTestSuite5%23teamcity%23org.apache.ignite.internal.processors.cache%23teamcity%23CacheSerializableTransactionsTest&pager.currentPage=1&order=DURATION_DESC&recordsPerPage=20&filterText=&status=&buildTypeId=IgniteTests24Java8_Cache5&buildId=2160072&tab=testsInfo]| and optimize them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10037) Cache 2 tests optimization
Alexey Platonov created IGNITE-10037: Summary: Cache 2 tests optimization Key: IGNITE-10037 URL: https://issues.apache.org/jira/browse/IGNITE-10037 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov We need to investigate how to optimize these tests: GridCachePartitionNotLoadedEventSelfTest.testPrimaryAndBackupDead CacheTxLoadingConcurrentGridStartSelfTestAllowOverwrite.testLoadCacheWithDataStreamerSequentialWithConfigAndRestarts IgniteCacheEntryProcessorNodeJoinTest.testEntryProcessorNodeLeave CacheTxLoadingConcurrentGridStartSelfTestAllowOverwrite.testLoadCacheWithDataStreamerSequentialWithConfig CacheTxLoadingConcurrentGridStartSelfTestAllowOverwrite.testLoadCacheWithDataStreamerSequentialClientWithConfig and optimize them. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-10008) Need to investigate why these tests is failed after consistentID set
Alexey Platonov created IGNITE-10008: Summary: Need to investigate why these tests is failed after consistentID set Key: IGNITE-10008 URL: https://issues.apache.org/jira/browse/IGNITE-10008 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov In these test method setConsistentID is overridden because setting of consistentID to nodes will let fail them: JdbcDistributedJoinsQueryTest GridCacheAbstractMetricsSelfTest IgniteCacheContinuousQueryNoUnsubscribeTest IgniteBinaryObjectFieldsQuerySelfTest IgniteSqlSkipReducerOnUpdateDmlSelfTest -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9979) [ML] Implement Naive Bayes classifier over Discrete Disctribution
Alexey Platonov created IGNITE-9979: --- Summary: [ML] Implement Naive Bayes classifier over Discrete Disctribution Key: IGNITE-9979 URL: https://issues.apache.org/jira/browse/IGNITE-9979 Project: Ignite Issue Type: Sub-task Components: ml Reporter: Alexey Platonov Fix For: 2.8 There is need to create Naive Bayes classifier over simple Discrete Distribution for non-continuous features in dataset. As naive implementation it may be set of histograms of feature values for each feature in dataset. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9978) [ML] Implement Compound Naive Bayes classifier
Alexey Platonov created IGNITE-9978: --- Summary: [ML] Implement Compound Naive Bayes classifier Key: IGNITE-9978 URL: https://issues.apache.org/jira/browse/IGNITE-9978 Project: Ignite Issue Type: Task Components: ml Reporter: Alexey Platonov Fix For: 2.8 We need to create compound Naive Bayes classifier as model composition of several Naive Bayes classifiers where each classifier represents subset of features of one type. For example such model may contain Naive Bayes model over Gauss Distribution for all continuous features and Naive Bayes model over Discrete Distribution for enum-like features. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9959) Set consistent id in tests
Alexey Platonov created IGNITE-9959: --- Summary: Set consistent id in tests Key: IGNITE-9959 URL: https://issues.apache.org/jira/browse/IGNITE-9959 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov We need to set consistent id for starting nodes in tests -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9930) Split Zookeeper Discovery 2 onto several configs
Alexey Platonov created IGNITE-9930: --- Summary: Split Zookeeper Discovery 2 onto several configs Key: IGNITE-9930 URL: https://issues.apache.org/jira/browse/IGNITE-9930 Project: Ignite Issue Type: Improvement Reporter: Alexey Platonov Assignee: Alexey Platonov At the current state this config require a lot of time for one test-run (>100min). We need to split this onto several configs with running time <= 60min. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9886) Split Continuous Query 1 onto several config
Alexey Platonov created IGNITE-9886: --- Summary: Split Continuous Query 1 onto several config Key: IGNITE-9886 URL: https://issues.apache.org/jira/browse/IGNITE-9886 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov Assignee: Alexey Platonov At the current state this config require a lot of time for one test-run (>100min). We need to split this onto several configs with running time <= 60min. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9834) ClientImpl shouldn't fail JVM with 130 error code after node validation errors
Alexey Platonov created IGNITE-9834: --- Summary: ClientImpl shouldn't fail JVM with 130 error code after node validation errors Key: IGNITE-9834 URL: https://issues.apache.org/jira/browse/IGNITE-9834 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov Assignee: Alexey Platonov In the current number of Ignite if client (tcp-client-disco-msg-worker) receive validation error message represented by TcpDiscoveryCheckFailedMessage then client fail with 130 error code. Semantically such behavior is invalid. Client must cancel own work in normal order. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9829) Add validation error distinction to TcpDiscoveryCheckFailedMessage
Alexey Platonov created IGNITE-9829: --- Summary: Add validation error distinction to TcpDiscoveryCheckFailedMessage Key: IGNITE-9829 URL: https://issues.apache.org/jira/browse/IGNITE-9829 Project: Ignite Issue Type: Improvement Components: messaging Reporter: Alexey Platonov Fix For: 3.0 There is no way to define validation error type during joining on client side after node validation on server side. TcpDiscoveryCheckFailedMessage just aggregates error messages to one string. For example: if there was auth error while joining then client receive TcpDiscoveryCheckFailedMessage instead of TcpDiscoveryAuthFailedMessage. This is due to server validation protocol supports just IgniteNodeValidationResult as string-wrapper without additional information about types of validation errors. Set of enum values representing validation errors may be good solution. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9726) GridCacheAbstractFailoverSelfTest may lock all suite on put/remove cache operations
Alexey Platonov created IGNITE-9726: --- Summary: GridCacheAbstractFailoverSelfTest may lock all suite on put/remove cache operations Key: IGNITE-9726 URL: https://issues.apache.org/jira/browse/IGNITE-9726 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov Assignee: Alexey Platonov -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9675) Deadlock on Ignite:active() and stopping grid simultaneously calling
Alexey Platonov created IGNITE-9675:
---
Summary: Deadlock on Ignite:active() and stopping grid
simultaneously calling
Key: IGNITE-9675
URL: https://issues.apache.org/jira/browse/IGNITE-9675
Project: Ignite
Issue Type: Bug
Reporter: Alexey Platonov
Assignee: Alexey Platonov
There is deadlock on Ignite:active() and stopping grid simultaneously calling
# Trying to stop client node.
{code:java}
"main-ScalaTest-running-VisorInProcDriverSpec" #1 prio=5 os_prio=0
tid=0x7f267800e800 nid=0x6574 sleeping[0x7f2681b7e000]
java.lang.Thread.State: TIMED_WAITING (sleeping)
at java.lang.Thread.sleep(Native Method)
at
org.apache.ignite.internal.util.GridSpinReadWriteLock.writeLock(GridSpinReadWriteLock.java:206)
at
org.apache.ignite.internal.processors.task.GridTaskProcessor.onKernalStop(GridTaskProcessor.java:190)
at org.apache.ignite.internal.IgniteKernal.stop0(IgniteKernal.java:2135)
at org.apache.ignite.internal.IgniteKernal.stop(IgniteKernal.java:2083)
at
org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.stop0(IgnitionEx.java:2590)
- locked <0x000797ecf7f8> (a
org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance)
at
org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.stop(IgnitionEx.java:2553)
at org.apache.ignite.internal.IgnitionEx.stop(IgnitionEx.java:374)
at org.apache.ignite.Ignition.stop(Ignition.java:225)
{code}
and
2. Execute task that tries to get ignite.active() state, which also executes
task
GridClusterStateProcessor.sendComputeCheckGlobalState(GridClusterStateProcessor.java:1086)
{code:java}
"mgmt-#2470" #2965 prio=5 os_prio=0 tid=0x7f23fc001000 nid=0x730b waiting
on condition [0x7f221f0ee000]
java.lang.Thread.State: WAITING (parking)
at sun.misc.Unsafe.park(Native Method)
at java.util.concurrent.locks.LockSupport.park(LockSupport.java:304)
at
org.apache.ignite.internal.util.future.GridFutureAdapter.get0(GridFutureAdapter.java:177)
at
org.apache.ignite.internal.util.future.GridFutureAdapter.get(GridFutureAdapter.java:140)
at
org.apache.ignite.internal.AsyncSupportAdapter.saveOrGet(AsyncSupportAdapter.java:112)
at
org.apache.ignite.internal.IgniteComputeImpl.call(IgniteComputeImpl.java:786)
at
org.apache.ignite.internal.processors.cluster.GridClusterStateProcessor.sendComputeCheckGlobalState(GridClusterStateProcessor.java:1086)
at
org.apache.ignite.internal.processors.cluster.GridClusterStateProcessor.publicApiActiveState(GridClusterStateProcessor.java:177)
at
org.apache.ignite.internal.cluster.IgniteClusterImpl.active(IgniteClusterImpl.java:300)
at
org.apache.ignite.internal.visor.node.VisorNodeDataCollectorTask.reduce(VisorNodeDataCollectorTask.java:75)
at
org.gridgain.grid.internal.visor.node.VisorGridGainNodeDataCollectorTask.reduce0(VisorGridGainNodeDataCollectorTask.java:39)
at
org.gridgain.grid.internal.visor.node.VisorGridGainNodeDataCollectorTask.reduce0(VisorGridGainNodeDataCollectorTask.java:26)
at
org.apache.ignite.internal.visor.VisorMultiNodeTask.reduce(VisorMultiNodeTask.java:139)
at
org.apache.ignite.internal.processors.task.GridTaskWorker$6.call(GridTaskWorker.java:1139)
at
org.apache.ignite.internal.util.IgniteUtils.wrapThreadLoader(IgniteUtils.java:6765)
at
org.apache.ignite.internal.processors.task.GridTaskWorker.reduce(GridTaskWorker.java:1137)
at
org.apache.ignite.internal.processors.task.GridTaskWorker.onResponse(GridTaskWorker.java:964)
at
org.apache.ignite.internal.processors.task.GridTaskProcessor.processJobExecuteResponse(GridTaskProcessor.java:1081)
at
org.apache.ignite.internal.processors.task.GridTaskProcessor$JobMessageListener.onMessage(GridTaskProcessor.java:1316)
at
org.apache.ignite.internal.managers.communication.GridIoManager.invokeListener(GridIoManager.java:1569)
at
org.apache.ignite.internal.managers.communication.GridIoManager.processRegularMessage0(GridIoManager.java:1197)
at
org.apache.ignite.internal.managers.communication.GridIoManager.access$4200(GridIoManager.java:127)
at
org.apache.ignite.internal.managers.communication.GridIoManager$9.run(GridIoManager.java:1093)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
{code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
[jira] [Created] (IGNITE-9635) Eternal initial partition map exchange
Alexey Platonov created IGNITE-9635: --- Summary: Eternal initial partition map exchange Key: IGNITE-9635 URL: https://issues.apache.org/jira/browse/IGNITE-9635 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov Sometimes test suites times out on test class GridCacheReplicatedNodeRestartSelfTest. Finally, some test from this class blocked on awaiting node starting due to eternal initial partition map exchange. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9631) Timeouts in ZookeeperDIscoverySpi test suite
Alexey Platonov created IGNITE-9631: --- Summary: Timeouts in ZookeeperDIscoverySpi test suite Key: IGNITE-9631 URL: https://issues.apache.org/jira/browse/IGNITE-9631 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov Assignee: Alexey Platonov Sometimes this test suite times out. Failed test examples: [https://ci.ignite.apache.org/viewLog.html?buildId=1862120&tab=buildResultsDiv&buildTypeId=IgniteTests24Java8_ZooKeeperDiscovery2] [https://ci.ignite.apache.org/viewLog.html?buildId=1839809&tab=buildResultsDiv&buildTypeId=IgniteTests24Java8_ZooKeeperDiscovery2] It seems that test class GridCachePartitionedNodeRestartTest times out and block all suite. We need to add timeouts into test for suite timeouts preventing. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9580) Fix exit code 137 in Query 1 Suite
Alexey Platonov created IGNITE-9580: --- Summary: Fix exit code 137 in Query 1 Suite Key: IGNITE-9580 URL: https://issues.apache.org/jira/browse/IGNITE-9580 Project: Ignite Issue Type: Bug Reporter: Alexey Platonov Assignee: Alexey Platonov -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Created] (IGNITE-9531) ZookeeperDiscovery testClientReconnect is flaky in master
Alexey Platonov created IGNITE-9531:
---
Summary: ZookeeperDiscovery testClientReconnect is flaky in master
Key: IGNITE-9531
URL: https://issues.apache.org/jira/browse/IGNITE-9531
Project: Ignite
Issue Type: Bug
Reporter: Alexey Platonov
Assignee: Alexey Platonov
Fix For: 2.8
The test IgniteClientReconnectCacheTest#testReconnectMultinode(LongHistory)
periodically fails with timeouts in master.
>From the logs I see that the hang is caused by one of the two assertion errors:
{code}
java.lang.AssertionError
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.checkClientsStatus(ZookeeperDiscoveryImpl.java:1345)
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.access$2300(ZookeeperDiscoveryImpl.java:108)
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl$CheckClientsStatusCallback.processResult0(ZookeeperDiscoveryImpl.java:4332)
at
org.apache.ignite.spi.discovery.zk.internal.ZkAbstractChildrenCallback.processResult(ZkAbstractChildrenCallback.java:42)
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperClient$ChildrenCallbackWrapper.processResult(ZookeeperClient.java:1132)
at
org.apache.zookeeper.ClientCnxn$EventThread.processEvent(ClientCnxn.java:590)
at org.apache.zookeeper.ClientCnxn$EventThread.run(ClientCnxn.java:498)
{code}
or
{code}
java.lang.AssertionError
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.checkClientsStatus(ZookeeperDiscoveryImpl.java:1388)
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.access$2300(ZookeeperDiscoveryImpl.java:108)
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl$CheckClientsStatusCallback.processResult0(ZookeeperDiscoveryImpl.java:4332)
at
org.apache.ignite.spi.discovery.zk.internal.ZkAbstractChildrenCallback.processResult(ZkAbstractChildrenCallback.java:42)
at
org.apache.ignite.spi.discovery.zk.internal.ZookeeperClient$ChildrenCallbackWrapper.processResult(ZookeeperClient.java:1132)
at
org.apache.zookeeper.ClientCnxn$EventThread.processEvent(ClientCnxn.java:590)
at org.apache.zookeeper.ClientCnxn$EventThread.run(ClientCnxn.java:498)
{code}
The test failure can be rarely reproduced locally (run repeatedly with CPU
stress enabled).
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
[jira] [Created] (IGNITE-9022) [ML] Implement class labels mapping for SVM binary classifier
Alexey Platonov created IGNITE-9022:
---
Summary: [ML] Implement class labels mapping for SVM binary
classifier
Key: IGNITE-9022
URL: https://issues.apache.org/jira/browse/IGNITE-9022
Project: Ignite
Issue Type: Bug
Components: ml
Reporter: Alexey Platonov
Assignee: Aleksey Zinoviev
Fix For: 2.7
We need to automatically compute mapping from user's labels to \{-1;1} for SVM.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
