We finally figured out what was "wrong":
There's a call to
EODatabaseContext._EOAssertSafeMultiThreadedAccess(); on line 528 of
ERXObjectStoreCoordinatorSynchronizer.java that call results in the
"*** EODatabaseContext: access without lock" message being logged.
The good news is that EODatabaseContext in question ends up being
locked right after that...
Code snippet from
ERXObjectStoreCoordinatorSynchronizer.processRemoteChange();
EODatabaseContext dbc =
ERXEOAccessUtilities.databaseContextForEntityNamed(osc,
gid.entityName());
EODatabase database = dbc.database();
EODatabaseContext._EOAssertSafeMultiThreadedAccess(dbc);
dbc.lock();
I supposed all that needs to be done here is to lock the
EODatabaseContext before calling _EOAssertSafeMultiThreadedAccess();
How would one go about submitting a patch?
Thanks,
Saad
On Oct 23, 2008, at 6:32 PM, Chuck Hill wrote:
On Oct 23, 2008, at 2:55 PM, Saad Laassel wrote:
This is a follow on to this issue, we are deploying as a servlet
using WO 5.4.3 we modified our Database Context Delegate so it now
extends ERXDatabaseContextDelegate, we are initializing the remote
synchronizer manually by calling:
ERXObjectStoreCoordinatorSynchronizer
.synchronizer().initializeRemoteSynchronizer();
this is happening very early on in our application.java right after
we receive "ModelAddedNotification", everything seems to behave as
it should e.g. cached objects are being refreshed on different
instances except we are getting the following exception on a
frequent basis:
[Thread-2] DEBUG NSLog - *** EODatabaseContext: access without lock
Our application is usually deployed in a multi-server, multi-
instance environment we just want to make sure this is not going to
create any locking issues... any help is greatly appreciated ...
I'd take that warning very seriously. If that is not a bogus
message, it is a serious problem.
Chuck
On Oct 22, 2008, at 10:56 AM, Dov Rosenberg wrote:
We converted over to Project Wonder to use the
EJGroupsSynchronizer – everything looks like it is working OK
except I see this exception throughout the log file
[Thread-2] DEBUG NSLog - *** EODatabaseContext: access without
lock!
I did a thread dump to see what was executing when this was
occurring – it looks like the problem is in the
ERXObjectStoreCoordinatorSynchronizer.
"Thread-2" prio=6 tid=0x62d92780 nid=0x40c in Object.wait()
[0x6339f000..0x6339fd9c]
at java.lang.Object.wait(Native Method)
- waiting on <0x0d132540> (a java.util.LinkedList)
at java.lang.Object.wait(Unknown Source)
at er.extensions.eof.ERXObjectStoreCoordinatorSynchronizer
$
ProcessChangesQueue.run(ERXObjectStoreCoordinatorSynchronizer.java:
572)
- locked <0x0d132540> (a java.util.LinkedList)
at java.lang.Thread.run(Unknown Source)
Any thoughts on this? I am a little nervous about adding lock()
statements into the Wonder code.
Thanks in advance for any feedback.
Dov Rosenberg
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/slaassel%40inquira.com
This email sent to [EMAIL PROTECTED]
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/chill%40global-village.net
This email sent to [EMAIL PROTECTED]
--
Chuck Hill Senior Consultant / VP Development
Practical WebObjects - for developers who want to increase their
overall knowledge of WebObjects or who are trying to solve specific
problems.
http://www.global-village.net/products/practical_webobjects
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com
This email sent to [EMAIL PROTECTED]
