Ошибка transaction not connected or was disconnected

The following is a representative snippet of my code where an unexpected, at least on my part, exception is being thrown at the transaction.Rollback() statement.

The exception is of type NHibernate.TransactionException and the message is «Transaction not connected, or was disconnected.» And the stack trace looks like this: NHibernate.Transaction.AdoTransaction.CheckNotZombied() at NHibernate.Transaction.AdoTransaction.Rollback().

IEmployeeService employeeService = new EmployeeService(session);
var people = ReadFromFile('c:temp.csv');

for (var person in people)
{
    ITransaction transaction = session.BeginTransaction();
    try
    {
        employeeService.Update(person);

        employeeService.CheckForRecursion();

        transaction.Commit();
    }
    catch(Exception exp)
    {
        if (!transaction.WasRolledBack)
            transaction.Rollback();
    }
}

The CheckForRecursion uses some SQL to look for any recursion introduced by the last update, if so I want to undo. When recursion has been introduced then an exception bubbles up from SQL, rightly so, and I catch it and attempt to rollback. That’s when I encounter the error.

I have wrapped the rollback in a try catch so the whole thing can carry on but I see the same exception on each subsequent iteration of the for loop.

Ideas? Is this pattern correct?

The following is a representative snippet of my code where an unexpected, at least on my part, exception is being thrown at the transaction.Rollback() statement.

The exception is of type NHibernate.TransactionException and the message is «Transaction not connected, or was disconnected.» And the stack trace looks like this: NHibernate.Transaction.AdoTransaction.CheckNotZombied() at NHibernate.Transaction.AdoTransaction.Rollback().

IEmployeeService employeeService = new EmployeeService(session);
var people = ReadFromFile('c:temp.csv');

for (var person in people)
{
    ITransaction transaction = session.BeginTransaction();
    try
    {
        employeeService.Update(person);

        employeeService.CheckForRecursion();

        transaction.Commit();
    }
    catch(Exception exp)
    {
        if (!transaction.WasRolledBack)
            transaction.Rollback();
    }
}

The CheckForRecursion uses some SQL to look for any recursion introduced by the last update, if so I want to undo. When recursion has been introduced then an exception bubbles up from SQL, rightly so, and I catch it and attempt to rollback. That’s when I encounter the error.

I have wrapped the rollback in a try catch so the whole thing can carry on but I see the same exception on each subsequent iteration of the for loop.

Ideas? Is this pattern correct?

The following is a representative snippet of my code where an unexpected, at least on my part, exception is being thrown at the transaction.Rollback() statement.

The exception is of type NHibernate.TransactionException and the message is «Transaction not connected, or was disconnected.» And the stack trace looks like this: NHibernate.Transaction.AdoTransaction.CheckNotZombied() at NHibernate.Transaction.AdoTransaction.Rollback().

IEmployeeService employeeService = new EmployeeService(session);
var people = ReadFromFile('c:temp.csv');

for (var person in people)
{
    ITransaction transaction = session.BeginTransaction();
    try
    {
        employeeService.Update(person);

        employeeService.CheckForRecursion();

        transaction.Commit();
    }
    catch(Exception exp)
    {
        if (!transaction.WasRolledBack)
            transaction.Rollback();
    }
}

The CheckForRecursion uses some SQL to look for any recursion introduced by the last update, if so I want to undo. When recursion has been introduced then an exception bubbles up from SQL, rightly so, and I catch it and attempt to rollback. That’s when I encounter the error.

I have wrapped the rollback in a try catch so the whole thing can carry on but I see the same exception on each subsequent iteration of the for loop.

Ideas? Is this pattern correct?

The following is a representative snippet of my code where an unexpected, at least on my part, exception is being thrown at the transaction.Rollback() statement.

The exception is of type NHibernate.TransactionException and the message is «Transaction not connected, or was disconnected.» And the stack trace looks like this: NHibernate.Transaction.AdoTransaction.CheckNotZombied() at NHibernate.Transaction.AdoTransaction.Rollback().

IEmployeeService employeeService = new EmployeeService(session);
var people = ReadFromFile('c:temp.csv');

for (var person in people)
{
    ITransaction transaction = session.BeginTransaction();
    try
    {
        employeeService.Update(person);

        employeeService.CheckForRecursion();

        transaction.Commit();
    }
    catch(Exception exp)
    {
        if (!transaction.WasRolledBack)
            transaction.Rollback();
    }
}

The CheckForRecursion uses some SQL to look for any recursion introduced by the last update, if so I want to undo. When recursion has been introduced then an exception bubbles up from SQL, rightly so, and I catch it and attempt to rollback. That’s when I encounter the error.

I have wrapped the rollback in a try catch so the whole thing can carry on but I see the same exception on each subsequent iteration of the for loop.

Ideas? Is this pattern correct?

Hey!

First this library is great. Thanks for all the time and effort.

I’m seeing a «Couldn’t rollback ADO.NET connection. Transaction not connected» from Quartz.Impl.AdoJobStore.JobStoreTX error preceded by «An error occurred while scanning for the next trigger to fire.» from Quartz.Core.ErrorLogger.

Usually they come in batches of two and three intermittently. This has only occurred after moving to a high availability sqlserver cluster.

From my googling it seems like the quartz library will retry after a set time. This is corroborated by the fact that my jobs are getting work their work done.

Ultimately I’d like to get them out of my error logs if possible.

I’m running non clustered using the following quartz configuration:

add key=»quartz.scheduler.instanceName» value=»RemoteEventAPIScheduler»
add key=»quartz.scheduler.instanceId» value=»AUTO»
add key=»quartz.threadPool.type» value=»Quartz.Simpl.SimpleThreadPool, Quartz»
add key=»quartz.threadPool.threadCount» value=»10″
add key=»quartz.threadPool.threadPriority» value=»Normal»
add key=»quartz.scheduler.exporter.type» value=»Quartz.Simpl.RemotingSchedulerExporter, Quartz»
add key=»quartz.scheduler.exporter.port» value=»555″
add key=»quartz.scheduler.exporter.bindName» value=»QuartzScheduler»
add key=»quartz.scheduler.exporter.channelType» value=»tcp»
add key=»quartz.scheduler.exporter.channelName» value=»httpQuartz»
add key=»quartz.scheduler.jobFactory.type» value=»Events_Jobs.Factories.FlowJobFactory, Events Jobs»
add key=»quartz.jobStore.misfireThreshold» value=»480000″
add key=»quartz.jobStore.type» value=»Quartz.Impl.AdoJobStore.JobStoreTX, Quartz»
add key=»quartz.jobStore.tablePrefix» value=»QRTZ_»
add key=»quartz.jobStore.driverDelegateType» value=»Quartz.Impl.AdoJobStore.SqlServerDelegate, Quartz»
add key=»quartz.jobStore.dataSource» value=»myDS»
add key=»quartz.dataSource.myDS.provider» value=»SqlServer-20″
add key=»quartz.jobStore.useProperties» value=»true»
add key=»quartz.jobStore.clustered» value=»false»

Any guidance would be super appreciated. Thanks again.

NHibernate.TransactionException: Transaction not connected, or was disconnected
   at NHibernate.Transaction.AdoTransaction.CheckNotZombied() in d:CSharpNHNHnhibernatesrcNHibernateTransactionAdoTransaction.cs:line 406
   at NHibernate.Transaction.AdoTransaction.Rollback() in d:CSharpNHNHnhibernatesrcNHibernateTransactionAdoTransaction.cs:line 240

3 ответа

Лучший ответ

The server principal "AppUser" is not able to access the database "AppDB" under the current security context.