[hibernate-issues] [Hibernate-JIRA] Updated: (HHH-3392) Query Cache entries are not distributable
Steve Ebersole (JIRA)
noreply at atlassian.com
Wed Sep 10 00:49:08 EDT 2008
[ http://opensource.atlassian.com/projects/hibernate/browse/HHH-3392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Steve Ebersole updated HHH-3392:
--------------------------------
Attachment: HHH-3392-trunk.patch
HHH-3392-Branch_3_3.patch
HHH-3392-Branch_3_3.patch is a patch against the current state of the 3.3 branch. This one does not remove the public constructors on "effective singletons", but does provide better hashCode() impls as well as recalculating the hashCode() of QueryKey on deserialization
HHH-3392-trunk.patch is a patch against the current state of trunk (3.4). This one does the same as the 3.3 patch except that it does remove the public constructors of the singletons.
This time I would really like if someone could actually verify that this fixes the issue. I cannot commit this as of this moment as I am preparing a 3.3 release at the moment. If I can get quick turn around on this confirmation (in the affirmative) I can include these changes in that release (3.3.1).
> Query Cache entries are not distributable
> -----------------------------------------
>
> Key: HHH-3392
> URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-3392
> Project: Hibernate3
> Issue Type: Bug
> Components: caching (L2)
> Affects Versions: 3.3.0.CR1
> Environment: Any
> Reporter: Greg Luck
> Attachments: HHH-3392-Branch_3_3.patch, HHH-3392-trunk.patch
>
>
> Steve
> It will correct the issue we saw.
> On 17/06/2008, at 1:50 AM, Steve Ebersole wrote:
> I'm asking for confirmation that said change does in fact correct the
> issue :)
> If it does I'll make the change in Hibernate...
> On Mon, 2008-06-16 at 17:42 +1000, Greg Luck wrote:
> Steve
> Thanks. Why don't you make the change in Hibernate?
> On 14/06/2008, at 2:40 PM, Steve Ebersole wrote:
> Considering that RootEntityResultTransformer is not performing an
> kind
> of data uniqueing (singleton, etc) then yes that would be an issue.
> There is nothing wrong with caching a query which uses a
> ResultTransformer per-se.
> As simple test would be to add an equals() override
> to RootEntityResultTransformer where the class of the operands are
> evaluated:
> public boolean equals(Object other) {
> // all RootEntityResultTransformers are considered equal
> return other != null
> && RootEntityResultTransformer.class.isInstance( other )
> }
> On Tue, 2008-06-10 at 14:35 +1000, Greg Luck wrote:
> Steve
> I have a question from a Hibernate/Ehcache user.
> When he replicates a cached Query it does not get hit on the
> second
> box.
> I can see what the trouble is: the key includes an object
> reference to
> a RootEntityResultTransformer:
> ; transformer:
> org.hibernate.transform.RootEntityResultTransformer at a725aa,
> value=[4968486262001664, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
> version=1,
> hitCount=0, CreationTime = 1213009341528, LastAccessTime = 0 ]
> So, this cannot work. How does this work for any distributed
> cache?
> Begin forwarded message:
> From: alvins <alvins1982 at hotmail.com>
> Date: 9 June 2008 9:26:38 PM
> To: ehcache-list at lists.sourceforge.net
> Subject: [ehcache-list] Distributed ehcache - queries
> I am having some problems with ehcache in distributed mode and
> queries. I
> have a simple spring+hibernate+ehcache webapp running on two
> tomcat
> servers.
> I have enabled second level caching and I can see my simple pojo
> gets
> replicated back and forth fine between the two servers.
> The query caching seems to work fine if I have a single server
> started. I
> can see that is using the cached query and not hitting the db.
> However what
> I am trying to test is if I have two boxes - box1 does query and
> populates
> cache which gets replicated to box2 - then box2 doesn't need to
> do
> query
> since its in cache.
> My problem is that the query from box1 is not recognized as
> being
> the same
> as on box2 - it seems to get cached multiple times. Below is a
> log
> snippet
> of the two objects which are in my query cache - there should
> only
> be one.
> The first debug line is showing a query object which was
> replicated
> over
> from another box - the second line is showing a query object
> which
> was
> populated locally. This should not happen since they are the
> same
> query.
> 21:03:39,763 DEBUG net.sf.ehcache.store.MemoryStore -
> findTopXCache:
> findTopXMemoryStore hit for sql: select this_.emp_id as
> emp1_2_0_,
> this_.emp_firstname as emp2_2_0_, this_.emp_surname as emp3_2_0_
> from
> Employee this_; parameters: ; first row: 0; max rows: 10;
> transformer:
> org.hibernate.transform.RootEntityResultTransformer at a725aa
> LOCAL: [ key = sql: select this_.emp_id as emp1_2_0_,
> this_.emp_firstname as
> emp2_2_0_, this_.emp_surname as emp3_2_0_ from Employee this_;
> parameters: ;
> first row: 0; max rows: 10; transformer:
> org.hibernate.transform.RootEntityResultTransformer at a725aa,
> value=[4968486262001664, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
> version=1,
> hitCount=0, CreationTime = 1213009341528, LastAccessTime = 0 ]
> 21:03:39,763 DEBUG net.sf.ehcache.store.MemoryStore -
> findTopXCache:
> findTopXMemoryStore hit for sql: select this_.emp_id as
> emp1_2_0_,
> this_.emp_firstname as emp2_2_0_, this_.emp_surname as emp3_2_0_
> from
> Employee this_; parameters: ; first row: 0; max rows: 10;
> transformer:
> org.hibernate.transform.RootEntityResultTransformer at aa11a6
> LOCAL: [ key = sql: select this_.emp_id as emp1_2_0_,
> this_.emp_firstname as
> emp2_2_0_, this_.emp_surname as emp3_2_0_ from Employee this_;
> parameters: ;
> first row: 0; max rows: 10; transformer:
> org.hibernate.transform.RootEntityResultTransformer at aa11a6,
> value=[4968486194896896, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
> version=1,
> hitCount=5, CreationTime = 1213009325003, LastAccessTime =
> 1213009415332 ]
> Does anybody have any ideas?
> --
> View this message in context:
> http://www.nabble.com/Distributed-ehcache---queries-tp17730904p17730904.html
> Sent from the Ehcache mailing list archive at Nabble.com.
> -------------------------------------------------------------------------
> Check out the new SourceForge.net Marketplace.
> It's the best place to buy or sell services for
> just about anything Open Source.
> http://sourceforge.net/services/buy/index.php
> _______________________________________________
> ehcache-list mailing list
> ehcache-list at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ehcache-list
> Regards
> Greg Luck
> web: http://gregluck.com
> skype: gregrluck
> yahoo: gregrluck
> mobile: +61 408 061 622
> Regards
> Greg Luck
> web: http://gregluck.com
> skype: gregrluck
> yahoo: gregrluck
> mobile: +61 408 061 622
> Regards
> Greg Luck
> web: http://gregluck.com
> skype: gregrluck
> yahoo: gregrluck
> mobile: +61 408 061 622
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
More information about the hibernate-issues
mailing list