[jira] Updated: (DERBY-4741) Make Derby work reliably in the presence of thread interrupts

["Dag H. Wanvik (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/DERBY-4741?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dag H. Wanvik updated DERBY-4741:
---------------------------------

    Attachment: derby-4741-nio-container-2.log
                derby-4741-nio-container-2.stat
                derby-4741-nio-container-2.diff

Uploading an experimental patch which upon seeing the container channel interrupted/closed, 
closes and reopens the container to allow completion of the I/O.

Using a modified Derby151Test, the trace on my box (OpenSolaris snv_148, Java 1.6) shows how
the RAFContainer4.java I/O code recovers. When an interrupt is detected (in the form of an interrupted channel),
the thread's interrupt flag is tucked away in a thread local variable for now, and the flag is reset, so the thread can continue and retry the I/O operation when the container has been resurrected. 

The idea is that the thread local variable might be checked "higher up" somewhere, where throwing an exception would not make the database go down.

During this investigation, I have found numerous other locations at which an interrupt will make Derby go down, though, so RAFContainer4.java (or in deed NIO) is not the only weak spot we have.
Running the test on Windows, I see Derby choke on trying to switch log files, cf the enclosed derby.log file "derby-4741-nio-container-2.log" due to seeing a ChannelClosedException on the log file (NIO channel.force).

> Make Derby work reliably in the presence of thread interrupts
> -------------------------------------------------------------
>
>                 Key: DERBY-4741
>                 URL: https://issues.apache.org/jira/browse/DERBY-4741
>             Project: Derby
>          Issue Type: Bug
>          Components: Store
>    Affects Versions: 10.2.1.6, 10.2.2.0, 10.3.1.4, 10.3.2.1, 10.3.3.0, 10.4.1.3, 10.4.2.0, 10.5.1.1, 10.5.2.0, 10.5.3.0, 10.6.1.0
>            Reporter: Dag H. Wanvik
>         Attachments: derby-4741-nio-container-2.diff, derby-4741-nio-container-2.log, derby-4741-nio-container-2.stat
>
>
> When not executing on a small device VM, Derby has been using the Java NIO classes java.nio.clannel.* for file io.
> If thread is interrupted while executing blocking IO operations in NIO, the ClosedByInterruptException will get thrown. Unfortunately, Derby isn't current architected to retry and complete such operations (before passing on the interrupt), so the Derby database can be left in an inconsistent state and we therefore have to return a database level error. This means the applications can no longer access the database without a shutdown and reboot including a recovery.
> It would be nice if Derby could somehow detect and finish IO operations underway when thread interrupts happen before passing the exception on to the application. Derby embedded is sometimes embedded in applications that use Thread.interrupt to stop threads.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.