jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dominique Pfister" <dominique.pfis...@day.com>
Subject Re: DatabaseJournal doesn't work with springmodules
Date Fri, 10 Aug 2007 08:41:14 GMT
Hi Mike,

Sorry for the long delay, it must have skipped my attention! I will
verify the behaviour you described with the test cases you provided.

Kind regards
Dominique

On 8/9/07, mike.wilson <mtw@fluent.com> wrote:
>
> Hi,
>
> Just bumping this thread to see if anyone has had a chance to look at the
> test classes I posted.  I'm still seeing double-inserts when using
> DatabaseJournal along with transactions.  I've confirmed this behavior with
> released Jackrabbit 1.3.1.  The test classes posted to this thread on 7/26
> should demonstrate the difference:  DatabaseJournalTestNoTX.java (no
> transactions, no double-insert problem) and DatabaseJournalTestTX.java
> (transactions, double-insert problem.)  Again, no problem using FileJournal
> with transactions on or off.
>
> Thanks!
> Mike Wilson
>
>
> mike.wilson wrote:
> >
> >
> > Dominique Pfister wrote:
> >>
> >> Hi Mike,
> >>
> >> Please follow this link to find the issue:
> >>
> >>   https://issues.apache.org/jira/browse/JCR-884
> >>
> >> It should have been fixed in JR 1.3 and I therefore consider it very
> >> strange that you still experience it! Does the MySQL log output
> >> indicate that the _same_ JDBC connection inserts two records with the
> >> same revision_id? Can you provide some simple test that causes this
> >> failure in order to reproduce it?
> >>
> >> Kind regards
> >> Dominique
> >>
> >
> > Hi Dominique,
> >
> > Thanks very much for the confirmation and link - I wrote a simple test
> > case, and using this, I traced the behavior I wrote about to using
> > DatabaseJournal in conjunction with transactions.  I am still seeing
> > "double inserts" from the same JDBC connection (2 records inserted with
> > same revision ID) when using transactions.   I do not get the
> > double-insert behavior in the normal test case without transactions.
> >
> > I have attached the two test cases (with and without transactions).
> > http://www.nabble.com/file/p11813679/DatabaseJournalTest.zip
> > DatabaseJournalTest.zip  Am I doing something wrong with the transactions?
> >
> > Thanks again,
> > Mike Wilson
> >
> > On 7/24/07, mike.wilson <mtw@fluent.com> wrote:
> >>
> >>
> >> Dominique Pfister wrote:
> >> >
> >> > Hi Rafał,
> >> >
> >> > a test case will probably not be necessary. Looks like the
> >> > DatabaseJournal does not save the returned revision_id along with the
> >> > record currently being created but rather as one of its instance
> >> > member variables and therefore incorrectly associates both records
> >> > with the same revision_id. The commit() call after having inserted
> >> > only the first of two records is another defect: apparently the
> >> > DatabaseJournal's lock is not truly reentrant. I will file a jira
> >> > issue for this.
> >> >
> >> > Thanks for reporting this bug!
> >> > Dominique
> >> >
> >>
> >> Greetings,
> >>
> >> I am experiencing what appear to be the same symptoms as reported by
> >> Rafal.
> >> Has a fix been made, or is there an open jira issue on this?  I could not
> >> find either, but could have missed it.  To summarize:
> >>
> >> FileJournal works perfectly
> >> DatabaseJournal (mysql/innodb/read-committed) fails with errors like:
> >>
> >> Caused by:
> >> com.mysql.jdbc.exceptions.MySQLIntegrityConstraintViolationException:
> >> Duplicate entry '46' for key 1
> >>         at com.mysql.jdbc.SQLError.createSQLException(SQLError.java:931)
> >>         at com.mysql.jdbc.MysqlIO.checkErrorPacket(MysqlIO.java:2870)
> >>         at com.mysql.jdbc.MysqlIO.sendCommand(MysqlIO.java:1573)
> >>         at
> >> com.mysql.jdbc.ServerPreparedStatement.serverExecute(ServerPreparedStatement.java:1160)
> >>         at
> >> com.mysql.jdbc.ServerPreparedStatement.executeInternal(ServerPreparedStatement.java:685)
> >>         at
> >> com.mysql.jdbc.PreparedStatement.execute(PreparedStatement.java:790)
> >>         at
> >> org.apache.jackrabbit.core.journal.DatabaseJournal.append(DatabaseJournal.java:293)
> >>
> >> I get the same result whenever I update a node, using any combination of
> >> the
> >> following:
> >> Jackrabbit-1.3 (release)
> >> Jackrabbit-1.3.1 (candidate release from
> >> http://people.apache.org/~jukka/jackrabbit/1.3.1)
> >> MySQL 5.0.27
> >> mysql-connector-5.0.3
> >> mysql-connector-5.0.7
> >>
> >> When I turn on logging in mysql, I see duplicate commits/inserts into the
> >> journal, the first containing the node data, the second containing some
> >> short (null?) value, both having the same JOURNAL_ID:
> >>
> >> 2 Query       SET autocommit=0
> >> 2 Query       update J_C_GLOBAL_REVISION set revision_id = revision_id +
> >> 1
> >> 2 Query       select revision_id from J_C_GLOBAL_REVISION
> >> 2 Query       select REVISION_ID, JOURNAL_ID, PRODUCER_ID, REVISION_DATA
> >> from J_C_JOURNAL where REVISION_ID > 44
> >> 4 Query       select 1 from DEMO_REFS where NODE_ID =
> >> '6b594e07-c26a-4f12-ba91-e7699f15df3e'
> >> 2 Query       SET autocommit=0
> >> 2 Query       update J_C_GLOBAL_REVISION set revision_id = revision_id +
> >> 1
> >> 2 Query       select revision_id from J_C_GLOBAL_REVISION
> >> 2 Query       select REVISION_ID, JOURNAL_ID, PRODUCER_ID, REVISION_DATA
> >> from J_C_JOURNAL where REVISION_ID > 44
> >> ...
> >> (SQL calls that update the node appear here)
> >> ...
> >> 4 Query       commit
> >> 2 Query       insert into J_C_JOURNAL(REVISION_ID, JOURNAL_ID,
> >> PRODUCER_ID,
> >> REVISION_DATA) values (46,'node3','JR',<lots of data for the node>)
> >> 2 Query       commit
> >> 2 Query       SET autocommit=1
> >> 2 Query       SET autocommit=1
> >> 2 Query       insert into J_C_JOURNAL(REVISION_ID, JOURNAL_ID,
> >> PRODUCER_ID,
> >> REVISION_DATA) values (46,'node3','JR',_binary'\0\0')
> >> 2 Query       SET autocommit=1
> >> 2 Query       SET autocommit=1
> >>
> >>
> >> Thanks!
> >> Mike Wilson
> >>
> >>
> >>
> >> On 3/28/07, Rafał Kwiecień <rafal@consol.pl> wrote:
> >> > Hi Dominique
> >> >
> >> > Dnia środa, 28 marca 2007 09:46, Dominique Pfister napisał:
> >> > > Hi Rafał,
> >> > >
> >> > > judging from the fact that the revision id is incremented twice and
> >> > > then applied to two different records, I ask myself whether these
> >> > > records are appended from the *same* thread in the *same* instance.
> >> > Yes - the same thread and the same instance of DatabaseJournal
> >> >
> >> > > When appending a record to the DatabaseJournal, the order of SQL
> >> calls
> >> > > is as follows:
> >> > >
> >> > > -- BEGIN TRANSACTION --
> >> > > UPDATE global_revision SET revision_id = revision_id + 1
> >> > > SELECT revision_id FROM global_revision
> >> > > INSERT INTO revision(revision_id, ...) VALUES (?,...)
> >> > > -- COMMIT --
> >> >
> >> > I my case it looks like:
> >> >
> >> > -- BEGIN TRANSACTION --
> >> > UPDATE global_revision SET revision_id = revision_id + 1
> >> > SELECT revision_id FROM global_revision
> >> > UPDATE global_revision SET revision_id = revision_id + 1
> >> > SELECT revision_id FROM global_revision
> >> > INSERT INTO revision(revision_id, ...) VALUES (?,...)
> >> > -- COMMIT --
> >> > INSERT INTO revision(revision_id, ...) VALUES (?,...)
> >> > -- COMMIT --
> >> >
> >> > >
> >> > > The first call is supposed to lock the global_revision table, the
> >> > > second will fetch the new revision_id and the last one will append
a
> >> > > record with that new revision_id. If two instances would get the same
> >> > > revision_id, something would be seriously strange with the database
> >> > > transaction isolation. Two different threads in the same instance
> >> > > block each other until one has actually finished appending the new
> >> > > record. Therefore, the only conceivable scenario are  reentrant
> >> > > updates made by the same thread.
> >> >
> >> > After debugging I can say more about the problem. When transaction is
> >> > committed, there are methods prepare() and commit() called on
> >> > TransactionContext object. TransactionContext.resources contains five
> >> > objects: XAWorkspace, XAItemStateManager, XALockManager,
> >> XAVersionManager,
> >> > XAWorkspace. First TransactionContext.prepare() is called - and for
> >> each
> >> > resource prepare(). DatabaseJournal.doLock() is called twice (by
> >> > XAItemStateManager and XAVersionManager). When
> >> TransactionContext.commit()
> >> > is
> >> > called, also method commit() for each resource is called. And
> >> > DatabaseJournal.append() is called twice (also by XAItemStateManager
> >> and
> >> > XAVersionManager).
> >> >
> >> > >
> >> > > Can you reproduce your situation with a simple test case? This would
> >> > > definitely help...
> >> >
> >> > Sorry, I don't have simple test case. You can download spring-modules
> >> > package
> >> > and run jcr sample. If you add cluster configuration to
> >> > jackrabbit-repo.xml,
> >> > you get error.
> >> >
> >> >
> >> > --
> >> > Rafał Kwiecień
> >> > ConSol* Consulting & Solutions Software Poland Sp. z o.o.
> >> > ul. Piastowska 44C, 30-070 Kraków
> >> > http://www.consol.pl/
> >> >
> >>
> >>
> >>
> >> --
> >> View this message in context:
> >> http://www.nabble.com/DatabaseJournal-doesn%27t-work-with-springmodules-tf3452619.html#a11765520
> >> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
> >>
> >>
> >
> >
>
>
>
> --
> View this message in context: http://www.nabble.com/DatabaseJournal-doesn%27t-work-with-springmodules-tf3452619.html#a12078871
> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
>
>
Mime
View raw message