Re: [WARNING] Unstable 'develop' branch
On 24.10.2016 15:39, Niclas Hedhman wrote: Thanks for the help, Stan... Really appreciated that you still watch what is going on here. You're welcome, Niclas! :) I do watch this, just not always enough time to post meaningful reply, or the topic is a bit out of scope (e.g. entities) for me after all these years. :) On Mon, Oct 24, 2016 at 8:37 PM, Niclas Hedhmanwrote: Ii have figured it out... I missed the "reconnect to the jdbc_test_db as a superuser". My ltree addition happened outside the database "jdbc_test_db"... On my way now. On Mon, Oct 24, 2016 at 8:30 PM, Niclas Hedhman wrote: If I go back to psql, I get this informative message postgres=# CREATE EXTENSION ltree; ERROR: extension "ltree" already exists On Mon, Oct 24, 2016 at 8:23 PM, Niclas Hedhman wrote: Caused by: org.postgresql.util.PSQLException: ERROR: type ltree does not exist Position: 55 at org.postgresql.core.v3.QueryExecutorImpl.receiveErrorResponse(QueryExecutorImpl.java:2453) at org.postgresql.core.v3.QueryExecutorImpl.processResults(QueryExecutorImpl.java:2153) at org.postgresql.core.v3.QueryExecutorImpl.execute(QueryExecutorImpl.java:286) Is there an easy way to test ltree presence with the psql command line? On Mon, Oct 24, 2016 at 8:20 PM, Niclas Hedhman wrote: Ah, I see... stmt.execute( this._vendor.toString( d .createTableDefinitionBuilder() .setTableScope( TableScope.LOCAL_TEMPORARY ) .setTableName( t.tableName( "ltree_test" ) ) .setCommitAction( PgSQLTableCommitAction.DROP ) .setTableContentsSource( d.createTableElementListBuilder() .addTableElement( d.createColumnDefinition( "test_column", dt.userDefined( "ltree" ) ) ) .createExpression() ).createExpression() ) ); is the code executed and ANY SQLException will cause my InternalError, and SQLException is not shown. Adding the root cause and running again. On Mon, Oct 24, 2016 at 8:16 PM, Stanislav Muhametsin < stanislav.muhamet...@zest.mail.kapsi.fi> wrote: On 24.10.2016 15:05, Niclas Hedhman wrote: I managed to install postgres locally and set up according to your instructions. I even needed to figure out that postgresql-contrib package was needed for the "ltree" to work (accepted inside psql command line). However, I now get an error in the testcases; java.lang.InternalError: It seems that your database doesn't have ltree as type. It is needed to store collections. Please refer to hopefully supplied instructions on how to add ltree type (hint: run pg_install_dir/share/contrib/ltree.sql script or command 'CREATE EXTENSION ltree;'). root@devdesk:~# su postgres postgres@devdesk:/root$ psql could not change directory to "/root": Permission denied psql (9.4.9) Type "help" for help. postgres=# CREATE EXTENSION ltree; CREATE EXTENSION postgres=# \q Any ideas?? Indexing-SQL uses some SQL command to test whether ltree extension is installed. I totally forgot what is the command, and I can not access Zest codebase right now. Can you see the root cause via debugger, the first exception that is thrown, most likely causes this. It *might* be that the exception is something else, but the Indexing-SQL *thinks* it is because of failing ltree-test. On Mon, Oct 24, 2016 at 7:46 PM, Paul Merlin wrote: So, I ran integration tests and everything is stable except indexing-sql. A lot of the query tests fail. Something about mapping Identity to String is incomplete. I've attached the tests result report to ZEST-180 and pointed to the test setup instruction there. Paul Merlin a écrit : Niclas, The Identity change is quite massive, and affects many tests, especially in the Entity Stores and Indexing/Query subsystems. The changes builds on my local Linux system, but I notice that Redis and Riak (others?) test suites are disabled, and I have not look into making them run locally. So there may still be issues, and I am not at all surprised if the CI will fail. And I need to go to sleep now, so it may take a couple of days for this to stabilize. Cheers All integration tests that depend on external services are skipped if they can't reach the corresponding external service. For most of them, simply running the service with the default configuration is enough (Memcached, Redis, Riak, MongoDB). The SQL EntityStore is always tested against embedded databases (Derby, H2, SQLite). If a MySQL or PostgreSQL service is available it will be tested against them too. The SQL Index/Query is only tested against PostgreSQL if available. PostgreSQL needs some special setup (user/pass and some extension for indexing, documented in the corresponding extension). The CI do not run any of these external services. Three years ago I did start creating a Docker image with almost all of these services (https://github.com/eskatos/zest-docker-testbed). It's
Re: [WARNING] Unstable 'develop' branch
Ii have figured it out... I missed the "reconnect to the jdbc_test_db as a superuser". My ltree addition happened outside the database "jdbc_test_db"... On my way now. On Mon, Oct 24, 2016 at 8:30 PM, Niclas Hedhmanwrote: > If I go back to psql, I get this informative message > > postgres=# CREATE EXTENSION ltree; > ERROR: extension "ltree" already exists > > > > On Mon, Oct 24, 2016 at 8:23 PM, Niclas Hedhman > wrote: > >> >> Caused by: org.postgresql.util.PSQLException: ERROR: type ltree >> does not exist >> Position: 55 >>at >> org.postgresql.core.v3.QueryExecutorImpl.receiveErrorResponse(QueryExecutorImpl.java:2453) >>at >> org.postgresql.core.v3.QueryExecutorImpl.processResults(QueryExecutorImpl.java:2153) >>at >> org.postgresql.core.v3.QueryExecutorImpl.execute(QueryExecutorImpl.java:286) >> >> Is there an easy way to test ltree presence with the psql command line? >> >> >> On Mon, Oct 24, 2016 at 8:20 PM, Niclas Hedhman >> wrote: >> >>> Ah, I see... >>> >>> stmt.execute( this._vendor.toString( d >>> .createTableDefinitionBuilder() >>> .setTableScope( TableScope.LOCAL_TEMPORARY ) >>> .setTableName( t.tableName( "ltree_test" ) ) >>> .setCommitAction( PgSQLTableCommitAction.DROP ) >>> .setTableContentsSource( >>> d.createTableElementListBuilder() >>> .addTableElement( >>> d.createColumnDefinition( "test_column", dt.userDefined( >>> "ltree" ) ) ) >>> .createExpression() ).createExpression() ) ); >>> >>> >>> is the code executed and ANY SQLException will cause my InternalError, >>> and SQLException is not shown. Adding the root cause and running again. >>> >>> >>> >>> On Mon, Oct 24, 2016 at 8:16 PM, Stanislav Muhametsin < >>> stanislav.muhamet...@zest.mail.kapsi.fi> wrote: >>> On 24.10.2016 15:05, Niclas Hedhman wrote: > I managed to install postgres locally and set up according to your > instructions. I even needed to figure out that postgresql-contrib > package > was needed for the "ltree" to work (accepted inside psql command line). > > > However, I now get an error in the testcases; > > type="java.lang.InternalError">java.lang.InternalError: It seems that > your database doesn't have ltree as type. It is needed to store > collections. Please refer to hopefully supplied instructions on how to > add ltree type (hint: run > pg_install_dir/share/contrib/ltree.sql script or command > 'CREATE EXTENSION ltree;'). > > > root@devdesk:~# su postgres > postgres@devdesk:/root$ psql > could not change directory to "/root": Permission denied > psql (9.4.9) > Type "help" for help. > > postgres=# CREATE EXTENSION ltree; > CREATE EXTENSION > postgres=# \q > > > Any ideas?? > Indexing-SQL uses some SQL command to test whether ltree extension is installed. I totally forgot what is the command, and I can not access Zest codebase right now. Can you see the root cause via debugger, the first exception that is thrown, most likely causes this. It *might* be that the exception is something else, but the Indexing-SQL *thinks* it is because of failing ltree-test. > > > On Mon, Oct 24, 2016 at 7:46 PM, Paul Merlin > wrote: > > So, I ran integration tests and everything is stable except >> indexing-sql. >> >> A lot of the query tests fail. >> Something about mapping Identity to String is incomplete. >> >> I've attached the tests result report to ZEST-180 and pointed to the >> test setup instruction there. >> >> >> Paul Merlin a écrit : >> >>> Niclas, >>> >>> The Identity change is quite massive, and affects many tests, >>> especially in >> >>> the Entity Stores and Indexing/Query subsystems. The changes builds on my local Linux system, but I notice that Redis and Riak (others?) test suites are disabled, and I have not look into making them run locally. So there may still be issues, and I am not at all surprised if the CI will fail. And I need to go to sleep now, so it may take a couple of days for this >>> to >> >>> stabilize. Cheers >>> All integration tests that depend on external services are skipped if >>> they can't reach the corresponding external service. >>> >>> For most of them, simply running the service with the default >>> configuration is enough (Memcached, Redis, Riak, MongoDB). >>> >>> The SQL EntityStore is always tested against embedded databases >>> (Derby, >>> H2, SQLite). If a MySQL or PostgreSQL service is available it will be >>> tested against them too. The SQL
Re: [WARNING] Unstable 'develop' branch
On 24.10.2016 15:30, Niclas Hedhman wrote: If I go back to psql, I get this informative message postgres=# CREATE EXTENSION ltree; ERROR: extension "ltree" already exists OK, that is weird... Maybe 'ltree' extension is per-user? Have you tried logging in with same user as the Java tests are logging in there, and trying to run that SQL command (the CREATE TABLE command) by hand? It *might* be that ltree extension has changed over the years and no longer exposes 'ltree' type directly, altho I doubt that a lot.
Re: [WARNING] Unstable 'develop' branch
If I go back to psql, I get this informative message postgres=# CREATE EXTENSION ltree; ERROR: extension "ltree" already exists On Mon, Oct 24, 2016 at 8:23 PM, Niclas Hedhmanwrote: > > Caused by: org.postgresql.util.PSQLException: ERROR: type ltree > does not exist > Position: 55 >at > org.postgresql.core.v3.QueryExecutorImpl.receiveErrorResponse(QueryExecutorImpl.java:2453) >at > org.postgresql.core.v3.QueryExecutorImpl.processResults(QueryExecutorImpl.java:2153) >at > org.postgresql.core.v3.QueryExecutorImpl.execute(QueryExecutorImpl.java:286) > > Is there an easy way to test ltree presence with the psql command line? > > > On Mon, Oct 24, 2016 at 8:20 PM, Niclas Hedhman > wrote: > >> Ah, I see... >> >> stmt.execute( this._vendor.toString( d >> .createTableDefinitionBuilder() >> .setTableScope( TableScope.LOCAL_TEMPORARY ) >> .setTableName( t.tableName( "ltree_test" ) ) >> .setCommitAction( PgSQLTableCommitAction.DROP ) >> .setTableContentsSource( >> d.createTableElementListBuilder() >> .addTableElement( >> d.createColumnDefinition( "test_column", dt.userDefined( "ltree" >> ) ) ) >> .createExpression() ).createExpression() ) ); >> >> >> is the code executed and ANY SQLException will cause my InternalError, >> and SQLException is not shown. Adding the root cause and running again. >> >> >> >> On Mon, Oct 24, 2016 at 8:16 PM, Stanislav Muhametsin < >> stanislav.muhamet...@zest.mail.kapsi.fi> wrote: >> >>> On 24.10.2016 15:05, Niclas Hedhman wrote: >>> I managed to install postgres locally and set up according to your instructions. I even needed to figure out that postgresql-contrib package was needed for the "ltree" to work (accepted inside psql command line). However, I now get an error in the testcases; >>> type="java.lang.InternalError">java.lang.InternalError: It seems that your database doesn't have ltree as type. It is needed to store collections. Please refer to hopefully supplied instructions on how to add ltree type (hint: run pg_install_dir/share/contrib/ltree.sql script or command 'CREATE EXTENSION ltree;'). root@devdesk:~# su postgres postgres@devdesk:/root$ psql could not change directory to "/root": Permission denied psql (9.4.9) Type "help" for help. postgres=# CREATE EXTENSION ltree; CREATE EXTENSION postgres=# \q Any ideas?? >>> >>> Indexing-SQL uses some SQL command to test whether ltree extension is >>> installed. >>> I totally forgot what is the command, and I can not access Zest codebase >>> right now. >>> >>> Can you see the root cause via debugger, the first exception that is >>> thrown, most likely causes this. >>> It *might* be that the exception is something else, but the Indexing-SQL >>> *thinks* it is because of failing ltree-test. >>> >>> >>> >>> >>> On Mon, Oct 24, 2016 at 7:46 PM, Paul Merlin wrote: So, I ran integration tests and everything is stable except > indexing-sql. > > A lot of the query tests fail. > Something about mapping Identity to String is incomplete. > > I've attached the tests result report to ZEST-180 and pointed to the > test setup instruction there. > > > Paul Merlin a écrit : > >> Niclas, >> >> The Identity change is quite massive, and affects many tests, >>> >> especially in > >> the Entity Stores and Indexing/Query subsystems. >>> >>> The changes builds on my local Linux system, but I notice that Redis >>> and >>> Riak (others?) test suites are disabled, and I have not look into >>> making >>> them run locally. So there may still be issues, and I am not at all >>> surprised if the CI will fail. >>> >>> And I need to go to sleep now, so it may take a couple of days for >>> this >>> >> to > >> stabilize. >>> >>> Cheers >>> >> All integration tests that depend on external services are skipped if >> they can't reach the corresponding external service. >> >> For most of them, simply running the service with the default >> configuration is enough (Memcached, Redis, Riak, MongoDB). >> >> The SQL EntityStore is always tested against embedded databases >> (Derby, >> H2, SQLite). If a MySQL or PostgreSQL service is available it will be >> tested against them too. The SQL Index/Query is only tested against >> PostgreSQL if available. PostgreSQL needs some special setup >> (user/pass >> and some extension for indexing, documented in the corresponding >> > extension). > >> The CI do not run any of these external services. >> >> Three years ago I did start creating a Docker image with almost all of >> these services
Re: [WARNING] Unstable 'develop' branch
Caused by: org.postgresql.util.PSQLException: ERROR: type ltree does not exist Position: 55 at org.postgresql.core.v3.QueryExecutorImpl.receiveErrorResponse(QueryExecutorImpl.java:2453) at org.postgresql.core.v3.QueryExecutorImpl.processResults(QueryExecutorImpl.java:2153) at org.postgresql.core.v3.QueryExecutorImpl.execute(QueryExecutorImpl.java:286) Is there an easy way to test ltree presence with the psql command line? On Mon, Oct 24, 2016 at 8:20 PM, Niclas Hedhmanwrote: > Ah, I see... > > stmt.execute( this._vendor.toString( d > .createTableDefinitionBuilder() > .setTableScope( TableScope.LOCAL_TEMPORARY ) > .setTableName( t.tableName( "ltree_test" ) ) > .setCommitAction( PgSQLTableCommitAction.DROP ) > .setTableContentsSource( > d.createTableElementListBuilder() > .addTableElement( > d.createColumnDefinition( "test_column", dt.userDefined( "ltree" > ) ) ) > .createExpression() ).createExpression() ) ); > > > is the code executed and ANY SQLException will cause my InternalError, and > SQLException is not shown. Adding the root cause and running again. > > > > On Mon, Oct 24, 2016 at 8:16 PM, Stanislav Muhametsin < > stanislav.muhamet...@zest.mail.kapsi.fi> wrote: > >> On 24.10.2016 15:05, Niclas Hedhman wrote: >> >>> I managed to install postgres locally and set up according to your >>> instructions. I even needed to figure out that postgresql-contrib package >>> was needed for the "ltree" to work (accepted inside psql command line). >>> >>> >>> However, I now get an error in the testcases; >>> >>> >> type="java.lang.InternalError">java.lang.InternalError: It seems that >>> your database doesn't have ltree as type. It is needed to store >>> collections. Please refer to hopefully supplied instructions on how to >>> add ltree type (hint: run >>> pg_install_dir/share/contrib/ltree.sql script or command >>> 'CREATE EXTENSION ltree;'). >>> >>> >>> root@devdesk:~# su postgres >>> postgres@devdesk:/root$ psql >>> could not change directory to "/root": Permission denied >>> psql (9.4.9) >>> Type "help" for help. >>> >>> postgres=# CREATE EXTENSION ltree; >>> CREATE EXTENSION >>> postgres=# \q >>> >>> >>> Any ideas?? >>> >> >> Indexing-SQL uses some SQL command to test whether ltree extension is >> installed. >> I totally forgot what is the command, and I can not access Zest codebase >> right now. >> >> Can you see the root cause via debugger, the first exception that is >> thrown, most likely causes this. >> It *might* be that the exception is something else, but the Indexing-SQL >> *thinks* it is because of failing ltree-test. >> >> >> >> >> >>> >>> >>> On Mon, Oct 24, 2016 at 7:46 PM, Paul Merlin wrote: >>> >>> So, I ran integration tests and everything is stable except indexing-sql. A lot of the query tests fail. Something about mapping Identity to String is incomplete. I've attached the tests result report to ZEST-180 and pointed to the test setup instruction there. Paul Merlin a écrit : > Niclas, > > The Identity change is quite massive, and affects many tests, >> > especially in > the Entity Stores and Indexing/Query subsystems. >> >> The changes builds on my local Linux system, but I notice that Redis >> and >> Riak (others?) test suites are disabled, and I have not look into >> making >> them run locally. So there may still be issues, and I am not at all >> surprised if the CI will fail. >> >> And I need to go to sleep now, so it may take a couple of days for >> this >> > to > stabilize. >> >> Cheers >> > All integration tests that depend on external services are skipped if > they can't reach the corresponding external service. > > For most of them, simply running the service with the default > configuration is enough (Memcached, Redis, Riak, MongoDB). > > The SQL EntityStore is always tested against embedded databases (Derby, > H2, SQLite). If a MySQL or PostgreSQL service is available it will be > tested against them too. The SQL Index/Query is only tested against > PostgreSQL if available. PostgreSQL needs some special setup (user/pass > and some extension for indexing, documented in the corresponding > extension). > The CI do not run any of these external services. > > Three years ago I did start creating a Docker image with almost all of > these services (https://github.com/eskatos/zest-docker-testbed). It's > completely outdated now. I do have a very limited connection these days > so I won't be able to update it soon. But I have all the services > locally so I'll try and run the integration tests to see if something > broke. > Ideally we should build such a Docker image during the build and run > the > test suite into it
Re: [WARNING] Unstable 'develop' branch
Ah, I see... stmt.execute( this._vendor.toString( d .createTableDefinitionBuilder() .setTableScope( TableScope.LOCAL_TEMPORARY ) .setTableName( t.tableName( "ltree_test" ) ) .setCommitAction( PgSQLTableCommitAction.DROP ) .setTableContentsSource( d.createTableElementListBuilder() .addTableElement( d.createColumnDefinition( "test_column", dt.userDefined( "ltree" ) ) ) .createExpression() ).createExpression() ) ); is the code executed and ANY SQLException will cause my InternalError, and SQLException is not shown. Adding the root cause and running again. On Mon, Oct 24, 2016 at 8:16 PM, Stanislav Muhametsin < stanislav.muhamet...@zest.mail.kapsi.fi> wrote: > On 24.10.2016 15:05, Niclas Hedhman wrote: > >> I managed to install postgres locally and set up according to your >> instructions. I even needed to figure out that postgresql-contrib package >> was needed for the "ltree" to work (accepted inside psql command line). >> >> >> However, I now get an error in the testcases; >> >> > type="java.lang.InternalError">java.lang.InternalError: It seems that >> your database doesn't have ltree as type. It is needed to store >> collections. Please refer to hopefully supplied instructions on how to >> add ltree type (hint: run >> pg_install_dir/share/contrib/ltree.sql script or command >> 'CREATE EXTENSION ltree;'). >> >> >> root@devdesk:~# su postgres >> postgres@devdesk:/root$ psql >> could not change directory to "/root": Permission denied >> psql (9.4.9) >> Type "help" for help. >> >> postgres=# CREATE EXTENSION ltree; >> CREATE EXTENSION >> postgres=# \q >> >> >> Any ideas?? >> > > Indexing-SQL uses some SQL command to test whether ltree extension is > installed. > I totally forgot what is the command, and I can not access Zest codebase > right now. > > Can you see the root cause via debugger, the first exception that is > thrown, most likely causes this. > It *might* be that the exception is something else, but the Indexing-SQL > *thinks* it is because of failing ltree-test. > > > > > >> >> >> On Mon, Oct 24, 2016 at 7:46 PM, Paul Merlinwrote: >> >> So, I ran integration tests and everything is stable except indexing-sql. >>> >>> A lot of the query tests fail. >>> Something about mapping Identity to String is incomplete. >>> >>> I've attached the tests result report to ZEST-180 and pointed to the >>> test setup instruction there. >>> >>> >>> Paul Merlin a écrit : >>> Niclas, The Identity change is quite massive, and affects many tests, > especially in >>> the Entity Stores and Indexing/Query subsystems. > > The changes builds on my local Linux system, but I notice that Redis > and > Riak (others?) test suites are disabled, and I have not look into > making > them run locally. So there may still be issues, and I am not at all > surprised if the CI will fail. > > And I need to go to sleep now, so it may take a couple of days for this > to >>> stabilize. > > Cheers > All integration tests that depend on external services are skipped if they can't reach the corresponding external service. For most of them, simply running the service with the default configuration is enough (Memcached, Redis, Riak, MongoDB). The SQL EntityStore is always tested against embedded databases (Derby, H2, SQLite). If a MySQL or PostgreSQL service is available it will be tested against them too. The SQL Index/Query is only tested against PostgreSQL if available. PostgreSQL needs some special setup (user/pass and some extension for indexing, documented in the corresponding >>> extension). >>> The CI do not run any of these external services. Three years ago I did start creating a Docker image with almost all of these services (https://github.com/eskatos/zest-docker-testbed). It's completely outdated now. I do have a very limited connection these days so I won't be able to update it soon. But I have all the services locally so I'll try and run the integration tests to see if something >>> broke. >>> Ideally we should build such a Docker image during the build and run the test suite into it on CI. >> >> > -- Niclas Hedhman, Software Developer http://zest.apache.org - New Energy for Java
Re: [WARNING] Unstable 'develop' branch
So, I ran integration tests and everything is stable except indexing-sql. A lot of the query tests fail. Something about mapping Identity to String is incomplete. I've attached the tests result report to ZEST-180 and pointed to the test setup instruction there. Paul Merlin a écrit : > Niclas, > >> The Identity change is quite massive, and affects many tests, especially in >> the Entity Stores and Indexing/Query subsystems. >> >> The changes builds on my local Linux system, but I notice that Redis and >> Riak (others?) test suites are disabled, and I have not look into making >> them run locally. So there may still be issues, and I am not at all >> surprised if the CI will fail. >> >> And I need to go to sleep now, so it may take a couple of days for this to >> stabilize. >> >> Cheers > > All integration tests that depend on external services are skipped if > they can't reach the corresponding external service. > > For most of them, simply running the service with the default > configuration is enough (Memcached, Redis, Riak, MongoDB). > > The SQL EntityStore is always tested against embedded databases (Derby, > H2, SQLite). If a MySQL or PostgreSQL service is available it will be > tested against them too. The SQL Index/Query is only tested against > PostgreSQL if available. PostgreSQL needs some special setup (user/pass > and some extension for indexing, documented in the corresponding extension). > > The CI do not run any of these external services. > > Three years ago I did start creating a Docker image with almost all of > these services (https://github.com/eskatos/zest-docker-testbed). It's > completely outdated now. I do have a very limited connection these days > so I won't be able to update it soon. But I have all the services > locally so I'll try and run the integration tests to see if something broke. > > Ideally we should build such a Docker image during the build and run the > test suite into it on CI. > > >
Re: [WARNING] Unstable 'develop' branch
It seems that something has gone wrong. The changes (the many commit emails confirms that) has not reached GitHub mirror, which is normally only seconds away. I will try to ping infra about it. Niclas On Sun, Oct 23, 2016 at 10:20 PM, Niclas Hedhmanwrote: > The Identity change is quite massive, and affects many tests, especially > in the Entity Stores and Indexing/Query subsystems. > > The changes builds on my local Linux system, but I notice that Redis and > Riak (others?) test suites are disabled, and I have not look into making > them run locally. So there may still be issues, and I am not at all > surprised if the CI will fail. > > And I need to go to sleep now, so it may take a couple of days for this to > stabilize. > > Cheers > -- > Niclas Hedhman, Software Developer > http://zest.apache.org - New Energy for Java > -- Niclas Hedhman, Software Developer http://zest.apache.org - New Energy for Java