Why JBOSS ORACLE datasource doesn't fail fast when query times out?
Environment: jboss-eap-7.0
I was waiting a fast fail with the exception:
java.sql.SQLTimeoutException: ORA-01013 user requested cancel of current operation
But I get the following after the end of transaction.
Caused by: java.lang.Throwable: setRollbackOnly called from:
at com.arjuna.ats.internal.jta.transaction.arjunacore.TransactionImple.setRollbackOnly(TransactionImple.java:339)
How do I force it to fail fast?
This is related to the JDBC driver version and firewall stuff.
This is not related to JBOSS, XA, JTA, DATASOURCE or something else.
See: https://www.ibm.com/mysupport/s/question/0D50z000062kF2p/why-is-the-webspheredefaultquerytimeout-property-for-timing-out-sql-queries-not-working
Related
SpringBoot's embeded tomcat throws a IllegalMonitorStateException when writing data back to some http request in our production environment. It happens more frequently at high load.
tomcat verion: tomcat-embed-core:9.0.31
Below is part of the stacktrace:
java.lang.IllegalMonitorStateException: null
at java.lang.Object.wait(Native Method)
at org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper.doWrite(NioEndpoint.java:1290)
at org.apache.tomcat.util.net.SocketWrapperBase.doWrite(SocketWrapperBase.java:741)
at org.apache.tomcat.util.net.SocketWrapperBase.flushBlocking(SocketWrapperBase.java:694)
at org.apache.tomcat.util.net.SocketWrapperBase.flush(SocketWrapperBase.java:684)
THANKS.
You can try to upgrade the version to 9.0.32, as the part of the logic that you reported an error on, was reverted in 9.0.32.
https://github.com/apache/tomcat/commit/6e60713c75141bc00f03f08f759df993a6416c71
I am using Eclipse to develope a webapp. Everything was working fine, but sometimes when I tried to publist->start my server, I received the following error:
Oct 03, 2019 6:52:55 PM org.apache.catalina.loader.WebappClassLoaderBase checkStateForResourceLoading
INFO: Illegal access: this web application instance has been stopped already. Could not load []. The following stack trace is thrown for debugging purposes as well as to attempt to terminate the thread which caused the illegal access.
java.lang.IllegalStateException: Illegal access: this web application instance has been stopped already. Could not load []. The following stack trace is thrown for debugging purposes as well as to attempt to terminate the thread which caused the illegal access.
at org.apache.catalina.loader.WebappClassLoaderBase.checkStateForResourceLoading(WebappClassLoaderBase.java:1383)
at org.apache.catalina.loader.WebappClassLoaderBase.getResource(WebappClassLoaderBase.java:1036)
at com.mysql.cj.jdbc.AbandonedConnectionCleanupThread.checkThreadContextClassLoader(AbandonedConnectionCleanupThread.java:117)
at com.mysql.cj.jdbc.AbandonedConnectionCleanupThread.run(AbandonedConnectionCleanupThread.java:84)
at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128)
at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628)
at java.base/java.lang.Thread.run(Thread.java:834)
For the first few times, I just restarted the server, and it was working fine, but now even a restart doesent solve the problem. I read about it online, and some sources mentioned that it might be a thread management issue, but I couldn't get further than that. Can anybody please provide some info? Thanks.
The mysql driver is trying to clean-up on web application shutdown which is good. Unfortunately it is trying to load additional classes to perform this clean-up and - since the web application is shutting down - class loading is no longer available. Ideally, the driver would load all the classes it needed to clean-up when it firsts loads - that would avoid this issue.
If you can figure out which classes it is trying to load - a check of the driver source code should show you which - then you should be able to load them yourself on application start in, for example, a ServletContextListener.
If you created any Entity class and forgot to give ID to any specific column then also this problem occurs. So assign #ID to any column and run again. your issue will resolve.
I using springboot with HikariCP, but after a while my app crash and I got the error:
org.springframework.transaction.CannotCreateTransactionException: Could not open JPA EntityManager for transaction; nested exception is org.hibernate.exception.JDBCConnectionException: Unable to acquire JDBC Connection
...
Caused by: org.hibernate.exception.JDBCConnectionException: Unable to acquire JDBC Connection
....
Caused by: java.sql.SQLTransientConnectionException: HikariPool-6 - Connection is not available, request timed out after 30000ms.
This is my aplication.properties
spring.datasource.type=com.zaxxer.hikari.HikariDataSource
spring.datasource.url=jdbc:postgresql://localhost:5432/db_dnaso
#spring.datasource.url=jdbc:postgresql://172.16.1.10:5432/db_dnaso
spring.datasource.username=postgres
spring.datasource.password=dna44100
spring.datasource.driver-class-name=org.postgresql.Driver
So I have a lot of save, find and anothers access to DB, how can I visualize how method are blocking my connection?
tks
It looks like your your database server is running out of connection. Default value for property maximumPoolSize in Hikari is 10. That means it will try to create 10 connection on server startup and it will not start if not able to acquire 10 connection or may fail if your db server pool size having less connection in the pool as you are creating using Hikari configuration.
If you are able to start Spring Boot server and then facing this issue then try enabling leakDetectionThreshold and check which connection is taking more time and not returning to Hikari pool .
spring:
datasource:
hikari:
leak-detection-threshold: 2000
Enable the leakDetectionThreshold, set to something like 1 minute (60000ms). It is likely that you have a connection leak somewhere ... a connection is borrowed but never closed (returned).
In springboot, you can set spring.datasource.hikari.leak-detection-threshold=10000 (in milliseconds) in application.properties.
And the default value for maximumPoolSize is 10. You can change it with spring.datasource.hikari.maximum-pool-size=xx.
I have an application presently running (without issue) on JBoss 6 that I am attempting to upgrade to run on WildFly 10.1. Much of this upgrade is going well. However, the upgrade from Hibernate 3.4 (on JBoss 6) to Hibernate 5.1 (on WildFly 10.1) is causing a few issues.
Specifically, in my persistence.xml, I include the following property.
<property name="hibernate.hbm2ddl.auto" value="update"/>
Please NOTE: I am NOT making any schema or other DB changes as part of the upgrade. Furthermore, I am pointing at the same database instance that has been successfully running under the JBoss 6/Hibernate 3.4 instance. Therefore, I am confident that inclusion of this property should have no actual work/update to do upon first run with the WildFly 10.1/Hibernate 5.1 version.
However, inclusion of this property appears to 1) erroneously determine that it needs to make updates and 2) fail to do so successfully. It results in the following stack trace:
Failed to start service jboss.persistenceunit."app.ear#PU": org.jboss.msc.service.StartException in service jboss.persistenceunit."PU": javax.persistence.PersistenceException: [PersistenceUnit: PU] Unable to build Hibernate SessionFactory
...
Caused by: javax.persistence.PersistenceException: [PersistenceUnit: PU] Unable to build Hibernate SessionFactory
...
Caused by: org.hibernate.tool.schema.spi.SchemaManagementException: Unable to execute schema management to JDBC target [create index company_id_index on APPROVER (COMPANY_ID)]
...
Caused by: org.postgresql.util.PSQLException: ERROR: relation "company_id_index" already exists
Again, the table and index in question already exist (as confirmed by the final error).
Is Hibernate now no longer case sensitive (COMPANY_ID_INDEX being different than company_id_index)?
If so, how can I configure it so that it is case insensitive as it used to be (Postgres defaults all of this to lower....)
TIA!
Doh! Face palm! I recently discovered that similar errors were also occurring related to hbm2ddl index creation with Hibernate 3.4/JBoss 6 as I am now experiencing with Hibernate 5.1/Wildfly 10.1; however, they were NOT preventing successful start up of the persistence module. Essentially, they were being only subtly suppressed. I'm not sure if this is an expected change related to the Hibernate versions or not, as they do prevent it's start up in Hibernate 5.1/Wildfly 10.1?
The underlying issue here turned out to be that index names must be unique across the entire schema in Postgres. So multiple entities each having a FK to a COMPANY_ID column must each have a unique name for the index. Indices are relations in Postgres (driving the unique across schema requirement).
Thank you for the suggestions and apologies for the confusion.
What events cause "org.springframework.transaction.TransactionSystemException - Could not roll back JPA transaction; nested exception is javax.persistence.PersistenceException: unexpected error when rollbacking" exceptions?
I got the same error with spring and Jpa.
The problem was with disconnection to Mysql database in case of timeout. I didn't use application during the night and in the morning I got the same error. When you inactive, spring do nothing and it get timeout.
To prevent it you can add into application.properties next lines:
spring.datasource.connection-test-query="SELECT 1"
spring.datasource.test-while-idle=true
spring.datasource.test-on-borrow=true
Now it works fine.
Also additional info from Virtustilus blog can be useful.
If the transaction is a XA-Transaction (distributed) then one of the participants might have rejected the rollback.