JRedisPipeline exception

[cjb]

When there is an exception thrown from the constructor of
JRedisPipeline it results in a thread leak. The constructor never
returns an object, however, a new thread is started. This thread
should be shutdown if the constructor fails. There is no way to do
it from the application, since there is no object.

[Joubin Houshyar]

Hi,

Please let me know what codebase you are using.  RC1 release, or from repo branches 1.2.n, master?

/R

--
You received this message because you are subscribed to the Google Groups "JRedis" group.
To post to this group, send email to jre...@googlegroups.com.
To unsubscribe from this group, send email to jredis+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/jredis?hl=en.

[carl bruecken]

I am using RC1.

initializeComponents will create the response-handle thread, but if connect fails the thread is leaked.


protected ConnectionBase (ConnectionSpec spec, boolean connectImmediately)
        throws ClientRuntimeException
    {
        try {
            this.spec = notNull(spec, "ConnectionSpec init parameter", ClientRuntimeException.class);
            socketAddress = new InetSocketAddress(spec.getAddress(), spec.getPort());
            initializeComponents();
            if(connectImmediately) {
                connect ();

[Joubin Houshyar]

Hi,

Did you see this: http://groups.google.com/group/jredis/browse_thread/thread/9b242ffbcc1ced49
Build from the 1.2.n branch (the likely RC2). Please let me know if
it is still an issue. It

/R

On May 26, 11:17 am, carl bruecken <brueck...@gmail.com> wrote:
> I am using RC1.
>
> initializeComponents will create the response-handle thread, but if connect
> fails the thread is leaked.
>
> protected ConnectionBase (ConnectionSpec spec, boolean connectImmediately)
>         throws ClientRuntimeException
>     {
>         try {
>             this.spec = notNull(spec, "ConnectionSpec init parameter",
> ClientRuntimeException.class);
>             socketAddress = new InetSocketAddress(spec.getAddress(),
> spec.getPort());
>             initializeComponents();
>             if(connectImmediately) {
>                 connect ();
>             }
>         }
>
>
>
> On Tue, May 25, 2010 at 5:29 PM, Joubin Houshyar <suno...@gmail.com> wrote:
> > Hi,
>
> > Please let me know what codebase you are using.  RC1 release, or from repo
> > branches 1.2.n, master?
>
> > /R
>

> > On Tue, May 25, 2010 at 4:13 PM, cjb <brueck...@gmail.com> wrote:
>
> >> When there is an exception thrown from the constructor of
> >> JRedisPipeline it results in a thread leak.   The constructor never
> >> returns an object, however, a new thread is started.   This thread
> >> should be shutdown if the constructor fails.   There is no way to do
> >> it from the application, since there is no object.
>
> >> --
> >> You received this message because you are subscribed to the Google Groups
> >> "JRedis" group.
> >> To post to this group, send email to jre...@googlegroups.com.
> >> To unsubscribe from this group, send email to

> >> jredis+un...@googlegroups.com<jredis%2Bunsu...@googlegroups.com>

> >> .
> >> For more options, visit this group at
> >>http://groups.google.com/group/jredis?hl=en.
>
> >  --
> > You received this message because you are subscribed to the Google Groups
> > "JRedis" group.
> > To post to this group, send email to jre...@googlegroups.com.
> > To unsubscribe from this group, send email to

> > jredis+un...@googlegroups.com<jredis%2Bunsu...@googlegroups.com>

[carl bruecken]

This is still an issue using Git 1.2.n branch.    The 'response-handler threads are still stuck on LinkedBlockingQueue.take().   I see there is an event listener, but it doesn't do anything in the case of a FAULT, even though it gets the FAULT event when the connection can't be established.

This is extremely easily to duplicate - by calling JRedisPipeline(connectionSpec) and passing connectionSpec that points to a server that isn't running.  Do this in a loop and you'll see that the threads are leaked and the constructor returns null;

Stack trace:  for 'response-handler' thread

sun.misc.Unsafe.park(Native Method)
java.util.concurrent.locks.LockSupport.park(LockSupport.java:158)
java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.await(AbstractQueuedSynchronizer.java:1925)
java.util.concurrent.LinkedBlockingQueue.take(LinkedBlockingQueue.java:399)
org.jredis.ri.alphazero.connection.PipelineConnectionBase$ResponseHandler.run(PipelineConnectionBase.java:287)
java.lang.Thread.run(Thread.java:637)


  public void onEvent (Event event) {
            if(event.getSource() != PipelineConnectionBase.this) {
                // BUG: what to do about it?
            }
            switch (event.getType()){
                case CONNECTED:
                    break;
                case DISCONNECTED:
                    break;
                case FAULTED:            
                    break;
           
            }

To unsubscribe from this group, send email to jredis+un...@googlegroups.com.