Re: Thin client test suites failure
FYI I found out how a Java version and addinional java module parameters are passed to a build after choosing jdk version in a "Run custom build" dialog. A version is controlled by a reverse.dep.*.env.JAVA_HOME configuration parameter specified to behave like a combobox in a "~[Obsolete] Run test suites" template configuration. Addinional module parameters are present as JVM_EXTRA_ARGS in the same template configuration. A value for it is taken from "~Build Apache Ignite~" (which runs before any test suite) in a form %dep.IgniteTests24Java8_BuildApacheIgnite.JVM_EXTRA_ARGS%, simpy speaking "~Build Apache Ignite~" JVM_EXTRA_ARGS value is borrowed. "~Build Apache Ignite~" configures that parameter in a build step "Setup additional arguments" by running a special command line script, args are chosen depending on Java version. пт, 28 июн. 2019 г. в 08:46, Павлухин Иван : > > After a little bit more experimentation I tend to think that adding > java.transaction module is not needed and even brings a negative > effect. E.g. if we use an ignite binary release and enable optional > jta module (by copying libraries to lib folder) then a node simply > fails to start (failing to resolve some needed classes). I will start > a separate thread related to JTA module. > > пт, 21 июн. 2019 г. в 14:46, Павлухин Иван : > > > > My next question is why do we need adding java.transaction module > > explicitly for Java 9 and 10? As I see since 11 version there is no > > such module and classes from it in modules bundled with JDK. So, I > > suppose either we do not need java.transaction module or something > > does not work with Java 11. It is complicated... > > > > чт, 20 июн. 2019 г. в 17:13, Павлухин Иван : > > > > > > Igniters, > > > > > > I found one thing which might be reason of an observed behavior. The > > > problem does not appear for maven-based tests, while e.g. Python thin > > > client tests use ignite.sh for starting a cluster. So, I compared JVM > > > launch options and observed a following extra option in maven-based > > > runs: > > > --patch-module=java.transaction=~/.m2/repository/javax/transaction/javax.transaction-api/1.3/javax.transaction-api-1.3.jar > > > > > > I missed when this trick was employed? Can anybody help with it? Does > > > it mean that ignite.sh does not work for all supported Java versions? > > > > > > чт, 20 июн. 2019 г. в 13:30, Павлухин Иван : > > > > > > > > By the way, what about Java 9? Do we run tests using it? > > > > > > > > чт, 20 июн. 2019 г. в 13:28, Dmitriy Pavlov : > > > > > > > > > > Actually, TC Bot is not actively developed. I would like to say, > > > > > moreover, > > > > > help is needed there. > > > > > > > > > > I agree that 10095 is (intentionally) vague. Will separate > > > > > notification > > > > > based on JDK version be supported or not, depends on the time > > > > > available to > > > > > complete this task. > > > > > > > > > > чт, 20 июн. 2019 г. в 12:02, Павлухин Иван : > > > > > > > > > > > Dmitriy, > > > > > > > > > > > > It is good to know that TC Bot is being developed actively. Just to > > > > > > double check. Will the mentioned issue [1] allow us to see alerts > > > > > > when > > > > > > a particular test fails only on a specific Java version? > > > > > > > > > > > > [1] https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > > > > > > > чт, 20 июн. 2019 г. в 10:26, Dmitriy Pavlov : > > > > > > > > > > > > > > Hi, > > > > > > > > > > > > > > About TC Bot: I was going to support filtering of builds from > > > > > > > same JVM > > > > > > for > > > > > > > TC Bot visa > > > > > > > https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > > > > > > > > > Now it is displayed only in UI as JDK8,JDK11 tags. I still do > > > > > > > investigations on how to retrieve/build run history faster than > > > > > > > it is > > > > > > now. > > > > > > > > > > > > > > Only nightly runs are random, and yes, Ignite is tested using > > > > > > > different > > > > > > VMs > > > > > > > because Ignite 2.7.5 declared support of, at least, Java 11. > > > > > > > > > > > > > > Sincerely, > > > > > > > Dmitriy Pavlov > > > > > > > > > > > > > > ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > > > > > > > > > > > > > > > Maxim, > > > > > > > > > > > > > > > > Well, it is an interesting point. First of all I want to > > > > > > > > believe that > > > > > > > > issues occurring only on a specific Java version are rare. If > > > > > > > > so, > > > > > > > > running tests on a single version for a visa should be enough. > > > > > > > > > > > > > > > > On the other hand, a uniform test environment sounds a good idea > > > > > > > > (especially for a visa), preventing a test status change by > > > > > > > > "magic". > > > > > > > > > > > > > > > > So, I think that it can be done as follows: > > > > > > > > 1. Run tests for a visa on a baseline Java version (11?). > > > > > > > > 2. Run Nightly builds on all supported versions, choosing > > > > > > > > version at
Re: Thin client test suites failure
After a little bit more experimentation I tend to think that adding java.transaction module is not needed and even brings a negative effect. E.g. if we use an ignite binary release and enable optional jta module (by copying libraries to lib folder) then a node simply fails to start (failing to resolve some needed classes). I will start a separate thread related to JTA module. пт, 21 июн. 2019 г. в 14:46, Павлухин Иван : > > My next question is why do we need adding java.transaction module > explicitly for Java 9 and 10? As I see since 11 version there is no > such module and classes from it in modules bundled with JDK. So, I > suppose either we do not need java.transaction module or something > does not work with Java 11. It is complicated... > > чт, 20 июн. 2019 г. в 17:13, Павлухин Иван : > > > > Igniters, > > > > I found one thing which might be reason of an observed behavior. The > > problem does not appear for maven-based tests, while e.g. Python thin > > client tests use ignite.sh for starting a cluster. So, I compared JVM > > launch options and observed a following extra option in maven-based > > runs: > > --patch-module=java.transaction=~/.m2/repository/javax/transaction/javax.transaction-api/1.3/javax.transaction-api-1.3.jar > > > > I missed when this trick was employed? Can anybody help with it? Does > > it mean that ignite.sh does not work for all supported Java versions? > > > > чт, 20 июн. 2019 г. в 13:30, Павлухин Иван : > > > > > > By the way, what about Java 9? Do we run tests using it? > > > > > > чт, 20 июн. 2019 г. в 13:28, Dmitriy Pavlov : > > > > > > > > Actually, TC Bot is not actively developed. I would like to say, > > > > moreover, > > > > help is needed there. > > > > > > > > I agree that 10095 is (intentionally) vague. Will separate notification > > > > based on JDK version be supported or not, depends on the time available > > > > to > > > > complete this task. > > > > > > > > чт, 20 июн. 2019 г. в 12:02, Павлухин Иван : > > > > > > > > > Dmitriy, > > > > > > > > > > It is good to know that TC Bot is being developed actively. Just to > > > > > double check. Will the mentioned issue [1] allow us to see alerts when > > > > > a particular test fails only on a specific Java version? > > > > > > > > > > [1] https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > > > > > чт, 20 июн. 2019 г. в 10:26, Dmitriy Pavlov : > > > > > > > > > > > > Hi, > > > > > > > > > > > > About TC Bot: I was going to support filtering of builds from same > > > > > > JVM > > > > > for > > > > > > TC Bot visa > > > > > > https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > > > > > > > Now it is displayed only in UI as JDK8,JDK11 tags. I still do > > > > > > investigations on how to retrieve/build run history faster than it > > > > > > is > > > > > now. > > > > > > > > > > > > Only nightly runs are random, and yes, Ignite is tested using > > > > > > different > > > > > VMs > > > > > > because Ignite 2.7.5 declared support of, at least, Java 11. > > > > > > > > > > > > Sincerely, > > > > > > Dmitriy Pavlov > > > > > > > > > > > > ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > > > > > > > > > > > > > Maxim, > > > > > > > > > > > > > > Well, it is an interesting point. First of all I want to believe > > > > > > > that > > > > > > > issues occurring only on a specific Java version are rare. If so, > > > > > > > running tests on a single version for a visa should be enough. > > > > > > > > > > > > > > On the other hand, a uniform test environment sounds a good idea > > > > > > > (especially for a visa), preventing a test status change by > > > > > > > "magic". > > > > > > > > > > > > > > So, I think that it can be done as follows: > > > > > > > 1. Run tests for a visa on a baseline Java version (11?). > > > > > > > 2. Run Nightly builds on all supported versions, choosing version > > > > > > > at > > > > > > > random is a one of options for that. (BTW, do we need some > > > > > > > changes in > > > > > > > TC bot to send a proper alert when test fails on some versions and > > > > > > > fails on others?). > > > > > > > > > > > > > > ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > > > > > > > > > > > > > > > Folks, > > > > > > > > > > > > > > > > Maybe I'm missing something, but what is the reason for setting > > > > > > > > JAVA_HOME randomly? For instance, some commit is working under > > > > > > > > jdk8 > > > > > > > > but fails under jdk9 (and we declare that 9-th version is > > > > > > > > supported > > > > > by > > > > > > > > Ignite). Should we merge this commit to the master branch (all > > > > > > > > test > > > > > > > > suites can be OK under jdk8)? > > > > > > > > > > > > > > > > I know, that running all tests under all java version leads us > > > > > > > > to > > > > > > > > Hell, but maybe we can have separated versions nightly builds? > > > > > > > > > > > > > > > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван > > > > > > > > > > > > > wrote: > > > > > > > > > > >
Re: Thin client test suites failure
My next question is why do we need adding java.transaction module explicitly for Java 9 and 10? As I see since 11 version there is no such module and classes from it in modules bundled with JDK. So, I suppose either we do not need java.transaction module or something does not work with Java 11. It is complicated... чт, 20 июн. 2019 г. в 17:13, Павлухин Иван : > > Igniters, > > I found one thing which might be reason of an observed behavior. The > problem does not appear for maven-based tests, while e.g. Python thin > client tests use ignite.sh for starting a cluster. So, I compared JVM > launch options and observed a following extra option in maven-based > runs: > --patch-module=java.transaction=~/.m2/repository/javax/transaction/javax.transaction-api/1.3/javax.transaction-api-1.3.jar > > I missed when this trick was employed? Can anybody help with it? Does > it mean that ignite.sh does not work for all supported Java versions? > > чт, 20 июн. 2019 г. в 13:30, Павлухин Иван : > > > > By the way, what about Java 9? Do we run tests using it? > > > > чт, 20 июн. 2019 г. в 13:28, Dmitriy Pavlov : > > > > > > Actually, TC Bot is not actively developed. I would like to say, moreover, > > > help is needed there. > > > > > > I agree that 10095 is (intentionally) vague. Will separate notification > > > based on JDK version be supported or not, depends on the time available to > > > complete this task. > > > > > > чт, 20 июн. 2019 г. в 12:02, Павлухин Иван : > > > > > > > Dmitriy, > > > > > > > > It is good to know that TC Bot is being developed actively. Just to > > > > double check. Will the mentioned issue [1] allow us to see alerts when > > > > a particular test fails only on a specific Java version? > > > > > > > > [1] https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > > > чт, 20 июн. 2019 г. в 10:26, Dmitriy Pavlov : > > > > > > > > > > Hi, > > > > > > > > > > About TC Bot: I was going to support filtering of builds from same JVM > > > > for > > > > > TC Bot visa > > > > > https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > > > > > Now it is displayed only in UI as JDK8,JDK11 tags. I still do > > > > > investigations on how to retrieve/build run history faster than it is > > > > now. > > > > > > > > > > Only nightly runs are random, and yes, Ignite is tested using > > > > > different > > > > VMs > > > > > because Ignite 2.7.5 declared support of, at least, Java 11. > > > > > > > > > > Sincerely, > > > > > Dmitriy Pavlov > > > > > > > > > > ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > > > > > > > > > > > Maxim, > > > > > > > > > > > > Well, it is an interesting point. First of all I want to believe > > > > > > that > > > > > > issues occurring only on a specific Java version are rare. If so, > > > > > > running tests on a single version for a visa should be enough. > > > > > > > > > > > > On the other hand, a uniform test environment sounds a good idea > > > > > > (especially for a visa), preventing a test status change by "magic". > > > > > > > > > > > > So, I think that it can be done as follows: > > > > > > 1. Run tests for a visa on a baseline Java version (11?). > > > > > > 2. Run Nightly builds on all supported versions, choosing version at > > > > > > random is a one of options for that. (BTW, do we need some changes > > > > > > in > > > > > > TC bot to send a proper alert when test fails on some versions and > > > > > > fails on others?). > > > > > > > > > > > > ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > > > > > > > > > > > > > Folks, > > > > > > > > > > > > > > Maybe I'm missing something, but what is the reason for setting > > > > > > > JAVA_HOME randomly? For instance, some commit is working under > > > > > > > jdk8 > > > > > > > but fails under jdk9 (and we declare that 9-th version is > > > > > > > supported > > > > by > > > > > > > Ignite). Should we merge this commit to the master branch (all > > > > > > > test > > > > > > > suites can be OK under jdk8)? > > > > > > > > > > > > > > I know, that running all tests under all java version leads us to > > > > > > > Hell, but maybe we can have separated versions nightly builds? > > > > > > > > > > > > > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван > > > > wrote: > > > > > > > > > > > > > > > > Dmitriy, > > > > > > > > > > > > > > > > Thank you for the hint! Java version seems to be the reason. I > > > > > > > > see > > > > > > > > that tests pass with Java 8 and 11 and fail with Java 10. See > > > > > > > > the > > > > > > > > exception below. Btw, do we test with Java 9? Have not seen that > > > > > > > > version in recent runs. > > > > > > > > java.lang.NoClassDefFoundError: > > > > > > > > javax/transaction/SystemException > > > > > > > > at java.base/java.lang.Class.forName0(Native Method) > > > > > > > > at java.base/java.lang.Class.forName(Class.java:291) > > > > > > > > at > > > > > > > > > > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > >
Re: Thin client test suites failure
Igniters, I found one thing which might be reason of an observed behavior. The problem does not appear for maven-based tests, while e.g. Python thin client tests use ignite.sh for starting a cluster. So, I compared JVM launch options and observed a following extra option in maven-based runs: --patch-module=java.transaction=~/.m2/repository/javax/transaction/javax.transaction-api/1.3/javax.transaction-api-1.3.jar I missed when this trick was employed? Can anybody help with it? Does it mean that ignite.sh does not work for all supported Java versions? чт, 20 июн. 2019 г. в 13:30, Павлухин Иван : > > By the way, what about Java 9? Do we run tests using it? > > чт, 20 июн. 2019 г. в 13:28, Dmitriy Pavlov : > > > > Actually, TC Bot is not actively developed. I would like to say, moreover, > > help is needed there. > > > > I agree that 10095 is (intentionally) vague. Will separate notification > > based on JDK version be supported or not, depends on the time available to > > complete this task. > > > > чт, 20 июн. 2019 г. в 12:02, Павлухин Иван : > > > > > Dmitriy, > > > > > > It is good to know that TC Bot is being developed actively. Just to > > > double check. Will the mentioned issue [1] allow us to see alerts when > > > a particular test fails only on a specific Java version? > > > > > > [1] https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > чт, 20 июн. 2019 г. в 10:26, Dmitriy Pavlov : > > > > > > > > Hi, > > > > > > > > About TC Bot: I was going to support filtering of builds from same JVM > > > for > > > > TC Bot visa > > > > https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > > > Now it is displayed only in UI as JDK8,JDK11 tags. I still do > > > > investigations on how to retrieve/build run history faster than it is > > > now. > > > > > > > > Only nightly runs are random, and yes, Ignite is tested using different > > > VMs > > > > because Ignite 2.7.5 declared support of, at least, Java 11. > > > > > > > > Sincerely, > > > > Dmitriy Pavlov > > > > > > > > ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > > > > > > > > > Maxim, > > > > > > > > > > Well, it is an interesting point. First of all I want to believe that > > > > > issues occurring only on a specific Java version are rare. If so, > > > > > running tests on a single version for a visa should be enough. > > > > > > > > > > On the other hand, a uniform test environment sounds a good idea > > > > > (especially for a visa), preventing a test status change by "magic". > > > > > > > > > > So, I think that it can be done as follows: > > > > > 1. Run tests for a visa on a baseline Java version (11?). > > > > > 2. Run Nightly builds on all supported versions, choosing version at > > > > > random is a one of options for that. (BTW, do we need some changes in > > > > > TC bot to send a proper alert when test fails on some versions and > > > > > fails on others?). > > > > > > > > > > ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > > > > > > > > > > > Folks, > > > > > > > > > > > > Maybe I'm missing something, but what is the reason for setting > > > > > > JAVA_HOME randomly? For instance, some commit is working under jdk8 > > > > > > but fails under jdk9 (and we declare that 9-th version is supported > > > by > > > > > > Ignite). Should we merge this commit to the master branch (all test > > > > > > suites can be OK under jdk8)? > > > > > > > > > > > > I know, that running all tests under all java version leads us to > > > > > > Hell, but maybe we can have separated versions nightly builds? > > > > > > > > > > > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван > > > wrote: > > > > > > > > > > > > > > Dmitriy, > > > > > > > > > > > > > > Thank you for the hint! Java version seems to be the reason. I see > > > > > > > that tests pass with Java 8 and 11 and fail with Java 10. See the > > > > > > > exception below. Btw, do we test with Java 9? Have not seen that > > > > > > > version in recent runs. > > > > > > > java.lang.NoClassDefFoundError: javax/transaction/SystemException > > > > > > > at java.base/java.lang.Class.forName0(Native Method) > > > > > > > at java.base/java.lang.Class.forName(Class.java:291) > > > > > > > at > > > > > > > > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > > > > > > > at > > > > > > > > org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) > > > > > > > at > > > > > > > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) > > > > > > > at > > > > > > > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) > > > > > > > at > > > > > > > > org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) > > > > > > > at > > > > > org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) > > > > > > > at > > > > > > > > org.apache.ignite.internal.Ignition
Re: Thin client test suites failure
By the way, what about Java 9? Do we run tests using it? чт, 20 июн. 2019 г. в 13:28, Dmitriy Pavlov : > > Actually, TC Bot is not actively developed. I would like to say, moreover, > help is needed there. > > I agree that 10095 is (intentionally) vague. Will separate notification > based on JDK version be supported or not, depends on the time available to > complete this task. > > чт, 20 июн. 2019 г. в 12:02, Павлухин Иван : > > > Dmitriy, > > > > It is good to know that TC Bot is being developed actively. Just to > > double check. Will the mentioned issue [1] allow us to see alerts when > > a particular test fails only on a specific Java version? > > > > [1] https://issues.apache.org/jira/browse/IGNITE-10095 > > > > чт, 20 июн. 2019 г. в 10:26, Dmitriy Pavlov : > > > > > > Hi, > > > > > > About TC Bot: I was going to support filtering of builds from same JVM > > for > > > TC Bot visa > > > https://issues.apache.org/jira/browse/IGNITE-10095 > > > > > > Now it is displayed only in UI as JDK8,JDK11 tags. I still do > > > investigations on how to retrieve/build run history faster than it is > > now. > > > > > > Only nightly runs are random, and yes, Ignite is tested using different > > VMs > > > because Ignite 2.7.5 declared support of, at least, Java 11. > > > > > > Sincerely, > > > Dmitriy Pavlov > > > > > > ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > > > > > > > Maxim, > > > > > > > > Well, it is an interesting point. First of all I want to believe that > > > > issues occurring only on a specific Java version are rare. If so, > > > > running tests on a single version for a visa should be enough. > > > > > > > > On the other hand, a uniform test environment sounds a good idea > > > > (especially for a visa), preventing a test status change by "magic". > > > > > > > > So, I think that it can be done as follows: > > > > 1. Run tests for a visa on a baseline Java version (11?). > > > > 2. Run Nightly builds on all supported versions, choosing version at > > > > random is a one of options for that. (BTW, do we need some changes in > > > > TC bot to send a proper alert when test fails on some versions and > > > > fails on others?). > > > > > > > > ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > > > > > > > > > Folks, > > > > > > > > > > Maybe I'm missing something, but what is the reason for setting > > > > > JAVA_HOME randomly? For instance, some commit is working under jdk8 > > > > > but fails under jdk9 (and we declare that 9-th version is supported > > by > > > > > Ignite). Should we merge this commit to the master branch (all test > > > > > suites can be OK under jdk8)? > > > > > > > > > > I know, that running all tests under all java version leads us to > > > > > Hell, but maybe we can have separated versions nightly builds? > > > > > > > > > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван > > wrote: > > > > > > > > > > > > Dmitriy, > > > > > > > > > > > > Thank you for the hint! Java version seems to be the reason. I see > > > > > > that tests pass with Java 8 and 11 and fail with Java 10. See the > > > > > > exception below. Btw, do we test with Java 9? Have not seen that > > > > > > version in recent runs. > > > > > > java.lang.NoClassDefFoundError: javax/transaction/SystemException > > > > > > at java.base/java.lang.Class.forName0(Native Method) > > > > > > at java.base/java.lang.Class.forName(Class.java:291) > > > > > > at > > > > > > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > > > > > > at > > > > > > org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) > > > > > > at > > > > > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) > > > > > > at > > > > > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) > > > > > > at > > > > > > org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) > > > > > > at > > > > org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) > > > > > > at > > > > > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1992) > > > > > > at > > > > > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1683) > > > > > > at > > > > org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1109) > > > > > > at > > > > > > org.apache.ignite.internal.IgnitionEx.startConfigurations(IgnitionEx.java:1027) > > > > > > at > > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:913) > > > > > > at > > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:812) > > > > > > at > > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:682) > > > > > > at > > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:651) > > > > > > at org.apache.ignite.Ignition.start(Ignition.java:346) > > > > > > at
Re: Thin client test suites failure
Actually, TC Bot is not actively developed. I would like to say, moreover, help is needed there. I agree that 10095 is (intentionally) vague. Will separate notification based on JDK version be supported or not, depends on the time available to complete this task. чт, 20 июн. 2019 г. в 12:02, Павлухин Иван : > Dmitriy, > > It is good to know that TC Bot is being developed actively. Just to > double check. Will the mentioned issue [1] allow us to see alerts when > a particular test fails only on a specific Java version? > > [1] https://issues.apache.org/jira/browse/IGNITE-10095 > > чт, 20 июн. 2019 г. в 10:26, Dmitriy Pavlov : > > > > Hi, > > > > About TC Bot: I was going to support filtering of builds from same JVM > for > > TC Bot visa > > https://issues.apache.org/jira/browse/IGNITE-10095 > > > > Now it is displayed only in UI as JDK8,JDK11 tags. I still do > > investigations on how to retrieve/build run history faster than it is > now. > > > > Only nightly runs are random, and yes, Ignite is tested using different > VMs > > because Ignite 2.7.5 declared support of, at least, Java 11. > > > > Sincerely, > > Dmitriy Pavlov > > > > ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > > > > > Maxim, > > > > > > Well, it is an interesting point. First of all I want to believe that > > > issues occurring only on a specific Java version are rare. If so, > > > running tests on a single version for a visa should be enough. > > > > > > On the other hand, a uniform test environment sounds a good idea > > > (especially for a visa), preventing a test status change by "magic". > > > > > > So, I think that it can be done as follows: > > > 1. Run tests for a visa on a baseline Java version (11?). > > > 2. Run Nightly builds on all supported versions, choosing version at > > > random is a one of options for that. (BTW, do we need some changes in > > > TC bot to send a proper alert when test fails on some versions and > > > fails on others?). > > > > > > ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > > > > > > > Folks, > > > > > > > > Maybe I'm missing something, but what is the reason for setting > > > > JAVA_HOME randomly? For instance, some commit is working under jdk8 > > > > but fails under jdk9 (and we declare that 9-th version is supported > by > > > > Ignite). Should we merge this commit to the master branch (all test > > > > suites can be OK under jdk8)? > > > > > > > > I know, that running all tests under all java version leads us to > > > > Hell, but maybe we can have separated versions nightly builds? > > > > > > > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван > wrote: > > > > > > > > > > Dmitriy, > > > > > > > > > > Thank you for the hint! Java version seems to be the reason. I see > > > > > that tests pass with Java 8 and 11 and fail with Java 10. See the > > > > > exception below. Btw, do we test with Java 9? Have not seen that > > > > > version in recent runs. > > > > > java.lang.NoClassDefFoundError: javax/transaction/SystemException > > > > > at java.base/java.lang.Class.forName0(Native Method) > > > > > at java.base/java.lang.Class.forName(Class.java:291) > > > > > at > > > > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > > > > > at > > > > org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) > > > > > at > > > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) > > > > > at > > > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) > > > > > at > > > > org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) > > > > > at > > > org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) > > > > > at > > > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1992) > > > > > at > > > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1683) > > > > > at > > > org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1109) > > > > > at > > > > org.apache.ignite.internal.IgnitionEx.startConfigurations(IgnitionEx.java:1027) > > > > > at > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:913) > > > > > at > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:812) > > > > > at > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:682) > > > > > at > org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:651) > > > > > at org.apache.ignite.Ignition.start(Ignition.java:346) > > > > > at > > > > org.apache.ignite.startup.cmdline.CommandLineStartup.main(CommandLineStartup.java:300) > > > > > Caused by: java.lang.ClassNotFoundException: > > > javax.transaction.SystemException > > > > > at > > > > java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:582) > > > > > at > > > > java.base/jd
Re: Thin client test suites failure
Dmitriy, It is good to know that TC Bot is being developed actively. Just to double check. Will the mentioned issue [1] allow us to see alerts when a particular test fails only on a specific Java version? [1] https://issues.apache.org/jira/browse/IGNITE-10095 чт, 20 июн. 2019 г. в 10:26, Dmitriy Pavlov : > > Hi, > > About TC Bot: I was going to support filtering of builds from same JVM for > TC Bot visa > https://issues.apache.org/jira/browse/IGNITE-10095 > > Now it is displayed only in UI as JDK8,JDK11 tags. I still do > investigations on how to retrieve/build run history faster than it is now. > > Only nightly runs are random, and yes, Ignite is tested using different VMs > because Ignite 2.7.5 declared support of, at least, Java 11. > > Sincerely, > Dmitriy Pavlov > > ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > > > Maxim, > > > > Well, it is an interesting point. First of all I want to believe that > > issues occurring only on a specific Java version are rare. If so, > > running tests on a single version for a visa should be enough. > > > > On the other hand, a uniform test environment sounds a good idea > > (especially for a visa), preventing a test status change by "magic". > > > > So, I think that it can be done as follows: > > 1. Run tests for a visa on a baseline Java version (11?). > > 2. Run Nightly builds on all supported versions, choosing version at > > random is a one of options for that. (BTW, do we need some changes in > > TC bot to send a proper alert when test fails on some versions and > > fails on others?). > > > > ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > > > > > Folks, > > > > > > Maybe I'm missing something, but what is the reason for setting > > > JAVA_HOME randomly? For instance, some commit is working under jdk8 > > > but fails under jdk9 (and we declare that 9-th version is supported by > > > Ignite). Should we merge this commit to the master branch (all test > > > suites can be OK under jdk8)? > > > > > > I know, that running all tests under all java version leads us to > > > Hell, but maybe we can have separated versions nightly builds? > > > > > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван wrote: > > > > > > > > Dmitriy, > > > > > > > > Thank you for the hint! Java version seems to be the reason. I see > > > > that tests pass with Java 8 and 11 and fail with Java 10. See the > > > > exception below. Btw, do we test with Java 9? Have not seen that > > > > version in recent runs. > > > > java.lang.NoClassDefFoundError: javax/transaction/SystemException > > > > at java.base/java.lang.Class.forName0(Native Method) > > > > at java.base/java.lang.Class.forName(Class.java:291) > > > > at > > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > > > > at > > org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) > > > > at > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) > > > > at > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) > > > > at > > org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) > > > > at > > org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) > > > > at > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1992) > > > > at > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1683) > > > > at > > org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1109) > > > > at > > org.apache.ignite.internal.IgnitionEx.startConfigurations(IgnitionEx.java:1027) > > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:913) > > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:812) > > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:682) > > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:651) > > > > at org.apache.ignite.Ignition.start(Ignition.java:346) > > > > at > > org.apache.ignite.startup.cmdline.CommandLineStartup.main(CommandLineStartup.java:300) > > > > Caused by: java.lang.ClassNotFoundException: > > javax.transaction.SystemException > > > > at > > java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:582) > > > > at > > java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:190) > > > > at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:499) > > > > ... 18 more > > > > > > > > вт, 18 июн. 2019 г. в 23:27, Dmitriy Pavlov : > > > > > > > > > > Hi Ivan, > > > > > > > > > > Can these failures be related to Java version? Java home is set > > randomly by > > > > > TC Bot. > > > > > > > > > > Dmitriy Pavlov > > > > > > > > > > вт, 18 июн. 2019 г. в 22:11, Павлухин Иван : > > > > > > > > > > > Hi igniters, > > > > > > > > > > > >
Re: Thin client test suites failure
Hi, About TC Bot: I was going to support filtering of builds from same JVM for TC Bot visa https://issues.apache.org/jira/browse/IGNITE-10095 Now it is displayed only in UI as JDK8,JDK11 tags. I still do investigations on how to retrieve/build run history faster than it is now. Only nightly runs are random, and yes, Ignite is tested using different VMs because Ignite 2.7.5 declared support of, at least, Java 11. Sincerely, Dmitriy Pavlov ср, 19 июн. 2019 г. в 22:13, Павлухин Иван : > Maxim, > > Well, it is an interesting point. First of all I want to believe that > issues occurring only on a specific Java version are rare. If so, > running tests on a single version for a visa should be enough. > > On the other hand, a uniform test environment sounds a good idea > (especially for a visa), preventing a test status change by "magic". > > So, I think that it can be done as follows: > 1. Run tests for a visa on a baseline Java version (11?). > 2. Run Nightly builds on all supported versions, choosing version at > random is a one of options for that. (BTW, do we need some changes in > TC bot to send a proper alert when test fails on some versions and > fails on others?). > > ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > > > Folks, > > > > Maybe I'm missing something, but what is the reason for setting > > JAVA_HOME randomly? For instance, some commit is working under jdk8 > > but fails under jdk9 (and we declare that 9-th version is supported by > > Ignite). Should we merge this commit to the master branch (all test > > suites can be OK under jdk8)? > > > > I know, that running all tests under all java version leads us to > > Hell, but maybe we can have separated versions nightly builds? > > > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван wrote: > > > > > > Dmitriy, > > > > > > Thank you for the hint! Java version seems to be the reason. I see > > > that tests pass with Java 8 and 11 and fail with Java 10. See the > > > exception below. Btw, do we test with Java 9? Have not seen that > > > version in recent runs. > > > java.lang.NoClassDefFoundError: javax/transaction/SystemException > > > at java.base/java.lang.Class.forName0(Native Method) > > > at java.base/java.lang.Class.forName(Class.java:291) > > > at > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > > > at > org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) > > > at > org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) > > > at > org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) > > > at > org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) > > > at > org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) > > > at > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1992) > > > at > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1683) > > > at > org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1109) > > > at > org.apache.ignite.internal.IgnitionEx.startConfigurations(IgnitionEx.java:1027) > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:913) > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:812) > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:682) > > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:651) > > > at org.apache.ignite.Ignition.start(Ignition.java:346) > > > at > org.apache.ignite.startup.cmdline.CommandLineStartup.main(CommandLineStartup.java:300) > > > Caused by: java.lang.ClassNotFoundException: > javax.transaction.SystemException > > > at > java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:582) > > > at > java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:190) > > > at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:499) > > > ... 18 more > > > > > > вт, 18 июн. 2019 г. в 23:27, Dmitriy Pavlov : > > > > > > > > Hi Ivan, > > > > > > > > Can these failures be related to Java version? Java home is set > randomly by > > > > TC Bot. > > > > > > > > Dmitriy Pavlov > > > > > > > > вт, 18 июн. 2019 г. в 22:11, Павлухин Иван : > > > > > > > > > Hi igniters, > > > > > > > > > > Does anyone know why python, php and nodejs [1, 2, 3] suites fail > so > > > > > frequently on TC? Is there any activity to deal with it? > > > > > > > > > > [1] > > > > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPython&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > > > [2] > > > > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientNodeJs&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > > > [3]
Re: Thin client test suites failure
Maxim, Well, it is an interesting point. First of all I want to believe that issues occurring only on a specific Java version are rare. If so, running tests on a single version for a visa should be enough. On the other hand, a uniform test environment sounds a good idea (especially for a visa), preventing a test status change by "magic". So, I think that it can be done as follows: 1. Run tests for a visa on a baseline Java version (11?). 2. Run Nightly builds on all supported versions, choosing version at random is a one of options for that. (BTW, do we need some changes in TC bot to send a proper alert when test fails on some versions and fails on others?). ср, 19 июн. 2019 г. в 20:33, Maxim Muzafarov : > > Folks, > > Maybe I'm missing something, but what is the reason for setting > JAVA_HOME randomly? For instance, some commit is working under jdk8 > but fails under jdk9 (and we declare that 9-th version is supported by > Ignite). Should we merge this commit to the master branch (all test > suites can be OK under jdk8)? > > I know, that running all tests under all java version leads us to > Hell, but maybe we can have separated versions nightly builds? > > On Wed, 19 Jun 2019 at 13:47, Павлухин Иван wrote: > > > > Dmitriy, > > > > Thank you for the hint! Java version seems to be the reason. I see > > that tests pass with Java 8 and 11 and fail with Java 10. See the > > exception below. Btw, do we test with Java 9? Have not seen that > > version in recent runs. > > java.lang.NoClassDefFoundError: javax/transaction/SystemException > > at java.base/java.lang.Class.forName0(Native Method) > > at java.base/java.lang.Class.forName(Class.java:291) > > at > > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > > at > > org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) > > at > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) > > at > > org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) > > at > > org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) > > at org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) > > at > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1992) > > at > > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1683) > > at org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1109) > > at > > org.apache.ignite.internal.IgnitionEx.startConfigurations(IgnitionEx.java:1027) > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:913) > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:812) > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:682) > > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:651) > > at org.apache.ignite.Ignition.start(Ignition.java:346) > > at > > org.apache.ignite.startup.cmdline.CommandLineStartup.main(CommandLineStartup.java:300) > > Caused by: java.lang.ClassNotFoundException: > > javax.transaction.SystemException > > at > > java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:582) > > at > > java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:190) > > at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:499) > > ... 18 more > > > > вт, 18 июн. 2019 г. в 23:27, Dmitriy Pavlov : > > > > > > Hi Ivan, > > > > > > Can these failures be related to Java version? Java home is set randomly > > > by > > > TC Bot. > > > > > > Dmitriy Pavlov > > > > > > вт, 18 июн. 2019 г. в 22:11, Павлухин Иван : > > > > > > > Hi igniters, > > > > > > > > Does anyone know why python, php and nodejs [1, 2, 3] suites fail so > > > > frequently on TC? Is there any activity to deal with it? > > > > > > > > [1] > > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPython&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > > [2] > > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientNodeJs&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > > [3] > > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPhp&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > > > > > > -- > > > > Best regards, > > > > Ivan Pavlukhin > > > > > > > > > > > > -- > > Best regards, > > Ivan Pavlukhin -- Best regards, Ivan Pavlukhin
Re: Thin client test suites failure
Folks, Maybe I'm missing something, but what is the reason for setting JAVA_HOME randomly? For instance, some commit is working under jdk8 but fails under jdk9 (and we declare that 9-th version is supported by Ignite). Should we merge this commit to the master branch (all test suites can be OK under jdk8)? I know, that running all tests under all java version leads us to Hell, but maybe we can have separated versions nightly builds? On Wed, 19 Jun 2019 at 13:47, Павлухин Иван wrote: > > Dmitriy, > > Thank you for the hint! Java version seems to be the reason. I see > that tests pass with Java 8 and 11 and fail with Java 10. See the > exception below. Btw, do we test with Java 9? Have not seen that > version in recent runs. > java.lang.NoClassDefFoundError: javax/transaction/SystemException > at java.base/java.lang.Class.forName0(Native Method) > at java.base/java.lang.Class.forName(Class.java:291) > at > org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) > at > org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) > at > org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) > at > org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) > at > org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) > at org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) > at > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1992) > at > org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1683) > at org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1109) > at > org.apache.ignite.internal.IgnitionEx.startConfigurations(IgnitionEx.java:1027) > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:913) > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:812) > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:682) > at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:651) > at org.apache.ignite.Ignition.start(Ignition.java:346) > at > org.apache.ignite.startup.cmdline.CommandLineStartup.main(CommandLineStartup.java:300) > Caused by: java.lang.ClassNotFoundException: javax.transaction.SystemException > at > java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:582) > at > java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:190) > at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:499) > ... 18 more > > вт, 18 июн. 2019 г. в 23:27, Dmitriy Pavlov : > > > > Hi Ivan, > > > > Can these failures be related to Java version? Java home is set randomly by > > TC Bot. > > > > Dmitriy Pavlov > > > > вт, 18 июн. 2019 г. в 22:11, Павлухин Иван : > > > > > Hi igniters, > > > > > > Does anyone know why python, php and nodejs [1, 2, 3] suites fail so > > > frequently on TC? Is there any activity to deal with it? > > > > > > [1] > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPython&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > [2] > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientNodeJs&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > [3] > > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPhp&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > > > > -- > > > Best regards, > > > Ivan Pavlukhin > > > > > > > -- > Best regards, > Ivan Pavlukhin
Re: Thin client test suites failure
Dmitriy, Thank you for the hint! Java version seems to be the reason. I see that tests pass with Java 8 and 11 and fail with Java 10. See the exception below. Btw, do we test with Java 9? Have not seen that version in recent runs. java.lang.NoClassDefFoundError: javax/transaction/SystemException at java.base/java.lang.Class.forName0(Native Method) at java.base/java.lang.Class.forName(Class.java:291) at org.apache.ignite.internal.IgniteComponentType.createOptional0(IgniteComponentType.java:249) at org.apache.ignite.internal.IgniteComponentType.createOptional(IgniteComponentType.java:235) at org.apache.ignite.internal.processors.cache.GridCacheProcessor.createSharedContext(GridCacheProcessor.java:3427) at org.apache.ignite.internal.processors.cache.GridCacheProcessor.start(GridCacheProcessor.java:869) at org.apache.ignite.internal.IgniteKernal.startProcessor(IgniteKernal.java:1886) at org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1135) at org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1992) at org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1683) at org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1109) at org.apache.ignite.internal.IgnitionEx.startConfigurations(IgnitionEx.java:1027) at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:913) at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:812) at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:682) at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:651) at org.apache.ignite.Ignition.start(Ignition.java:346) at org.apache.ignite.startup.cmdline.CommandLineStartup.main(CommandLineStartup.java:300) Caused by: java.lang.ClassNotFoundException: javax.transaction.SystemException at java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:582) at java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:190) at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:499) ... 18 more вт, 18 июн. 2019 г. в 23:27, Dmitriy Pavlov : > > Hi Ivan, > > Can these failures be related to Java version? Java home is set randomly by > TC Bot. > > Dmitriy Pavlov > > вт, 18 июн. 2019 г. в 22:11, Павлухин Иван : > > > Hi igniters, > > > > Does anyone know why python, php and nodejs [1, 2, 3] suites fail so > > frequently on TC? Is there any activity to deal with it? > > > > [1] > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPython&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > [2] > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientNodeJs&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > [3] > > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPhp&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > > > -- > > Best regards, > > Ivan Pavlukhin > > -- Best regards, Ivan Pavlukhin
Re: Thin client test suites failure
Hi Ivan, Can these failures be related to Java version? Java home is set randomly by TC Bot. Dmitriy Pavlov вт, 18 июн. 2019 г. в 22:11, Павлухин Иван : > Hi igniters, > > Does anyone know why python, php and nodejs [1, 2, 3] suites fail so > frequently on TC? Is there any activity to deal with it? > > [1] > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPython&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > [2] > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientNodeJs&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > [3] > https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_ThinClientPhp&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv > > -- > Best regards, > Ivan Pavlukhin >