squirrel-sql-develop Mailing List for SQuirreL SQL Client (Page 30)

Hi !

I made some additional pages for Oracle plugin. How can I include it to
official plugin's build ?

WBR, Alex.

Gollapudi, Amit C. (STL) wrote:

>Thanks for the patch.  I'll try it out sometime today.  I wonder if your
>test against Oracle 10g didn't fail because it was running locally.  
>  
>
My test which reproduced the error was against a local Oracle server (9i 
instead of 10g though)
While I'm here, I might add that looking at SchemaInfo I was happy to 
find that it only queries
getColumns once per table - after that it uses the results of the 
previous call.  So it definitely seems
to be limiting as best it can the number of calls to the database for 
meta info about tables.

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

Thanks for the patch.  I'll try it out sometime today.  I wonder if your
test against Oracle 10g didn't fail because it was running locally.  It
could, plausibly, have some correlation to the amount of time it takes
to grab the DatabaseMetaData from a remote server.

Amit

-----Original Message-----
From: Gerd Wagner [mailto:bir...@t-...]=20
Sent: Sunday, October 09, 2005 7:46 AM
To: Gollapudi, Amit C. (STL)
Cc: Rob...@co...;
squ...@li...
Subject: Re: [Squirrel-sql-develop] DatabaseMetaData & Thread Safety


Hi Amit and Rob,

since the DatabaseMetaData object is private in SchemaInfo and is used=20
only in two methods it's easy to synchronize the object. The attached=20
code introduces the necessary synchronized blocks.

Amit, it would be nice if you could try out the code and tell us about=20
the results. Please note that you need the newest version from CVS to=20
take the attached class as it is. I've just committed some changes to it

that deal with about the opposite of your problem (see RFE #1281766). If

you don't want to do a fresh update just look for "synchronized" in the=20
attached class. It should be easy to introduce my changes to the version

you have.

By the way, on Friday I managed to install Oracle 10g on my box :-). I=20
couldn't get DatabaseFailureReplicator to fail even with a hundred=20
threads. Probably I've got to few tables.

Gerd




Gollapudi, Amit C. (STL) wrote:
> Indeed.  The stream has already been closed message is usually=20
> followed by the other ones I mentioned in our tests.  Let me know if=20
> you need any help in diagnosing or fixing this problem.  I would be=20
> glad to see my favorite SQL tool become fully useable again.
>=20
> Amit
>=20
> -----Original Message-----
> From: Rob Manning [mailto:Rob...@co...]
> Sent: Friday, October 07, 2005 4:10 PM
> To: Gollapudi, Amit C. (STL)
> Cc: squ...@li...
> Subject: Re: [Squirrel-sql-develop] DatabaseMetaData & Thread Safety
>=20
>=20
> Gollapudi, Amit C. (STL) wrote:
>=20
>=20
>>Robert:
>>
>>Thank you for your quick response.  I'm attaching a standalone test
>>that we've developed to replicate the protocol violation messages we=20
>>see in Squirrel sessions.  It uses pure JDBC and a set of threads=20
>>attempting to share a DatabaseMetaData reference.  It fails, almost=20
>>invariably, irrespective of which JDK and Oracle JDBC driver versions=20
>>we try (including the 10g version).
>>
>>Funny thing is, we do have a DBA who is trying to help us analyze the
>>problem.  As I mentioned, it's been an arduous task to replicate the=20
>>problem.  However, I wish to clarify myself.  The problem with our=20
>>database going down doesn't have a confirmed direct correlation to the

>>problem we are discussing here.
>>
>>However, I am still interested in finding out if Squirrel is using the
>>JDBC API in a way that will generate unpredictable results.  As=20
>>developers use Squirrel SQL with the code completion plugin against=20
>>Oracle database instances, we get SQLExceptions with messages
"Protocol
>=20
>=20
>>Violation", "Maximum data type length exceeded", and "Connection is
>>closed" pretty regularly as we work with the SQL editor.
>>=20
>>
>=20
> This particular test gives me the following:
>=20
> java.sql.SQLException: Stream has already been closed
>     at
oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:134)
>     at
oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:179)
>     at
oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:269)
>     at oracle.jdbc.ttc7.LongTTCItem.getChars(LongTTCItem.java:189)
>     at
>
oracle.jdbc.dbaccess.DBDataSetImpl.getCharsStreamItem(DBDataSetImpl.java
> :1630)
>     at=20
>
oracle.jdbc.driver.OracleStatement.getCharsInternal(OracleStatement.java
> :3347)
>     at=20
>
oracle.jdbc.driver.OracleStatement.getStringValue(OracleStatement.java:3
> 556)
>     at=20
>
oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5
> 659)
>     at=20
>
oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5
> 622)
>     at=20
>
oracle.jdbc.driver.OracleResultSetImpl.getObject(OracleResultSetImpl.jav
> a:739)
>     at=20
>
DatabaseFailureReplicator$MetaDataTestCase.dumpResultSet(DatabaseFailure
> Replicator.java:69)
>     at=20
>
DatabaseFailureReplicator$MetaDataTestCase.run(DatabaseFailureReplicator
> .java:50)
>     at java.lang.Thread.run(Thread.java:534)
>=20
> What I find peculiar is that the getColumns implementation (from the
> driver I decompiled) creates a
> new PreparedStatement, executes it and returns the ResultSet, which=20
> should be independent of other
> ResultSet and their associated PreparedStatements.  In other words,=20
> closing one in thread A shouldn't
> have any effect on thread B's attempt to read another ResultSet. =20
> Definitely requires a bit more study.
>=20
> Rob
>=20
>=20
> CollabraSpace - Revolutionary Collaboration
> Visit us at http://www.collabraspace.com
> This message has been scanned for viruses by
> ClamAV v0.83
>=20
>=20
> ******* Confidentiality Notice *******
> This email, its electronic document attachments, and the contents of=20
> its website linkages may contain confidential health information. =20
> This information is intended solely for use by the individual or=20
> entity to whom it is addressed.  If you have received this information

> in error, please notify the sender immediately and arrange for the=20
> prompt destruction of the material and any accompanying attachments.
>=20
>=20
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Power Architecture Resource Center: Free content, downloads,=20
> discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl
> _______________________________________________
> Squirrel-sql-develop mailing list=20
> Squ...@li...
> https://lists.sourceforge.net/lists/listinfo/squirrel-sql-develop
>=20

Hi Amit and Rob,

since the DatabaseMetaData object is private in SchemaInfo and is used 
only in two methods it's easy to synchronize the object. The attached 
code introduces the necessary synchronized blocks.

Amit, it would be nice if you could try out the code and tell us about 
the results. Please note that you need the newest version from CVS to 
take the attached class as it is. I've just committed some changes to it 
that deal with about the opposite of your problem (see RFE #1281766). If 
you don't want to do a fresh update just look for "synchronized" in the 
attached class. It should be easy to introduce my changes to the version 
you have.

By the way, on Friday I managed to install Oracle 10g on my box :-). I 
couldn't get DatabaseFailureReplicator to fail even with a hundred 
threads. Probably I've got to few tables.

Gerd




Gollapudi, Amit C. (STL) wrote:
> Indeed.  The stream has already been closed message is usually followed
> by the other ones I mentioned in our tests.  Let me know if you need any
> help in diagnosing or fixing this problem.  I would be glad to see my
> favorite SQL tool become fully useable again.
> 
> Amit
> 
> -----Original Message-----
> From: Rob Manning [mailto:Rob...@co...] 
> Sent: Friday, October 07, 2005 4:10 PM
> To: Gollapudi, Amit C. (STL)
> Cc: squ...@li...
> Subject: Re: [Squirrel-sql-develop] DatabaseMetaData & Thread Safety
> 
> 
> Gollapudi, Amit C. (STL) wrote:
> 
> 
>>Robert:
>>
>>Thank you for your quick response.  I'm attaching a standalone test 
>>that we've developed to replicate the protocol violation messages we 
>>see in Squirrel sessions.  It uses pure JDBC and a set of threads 
>>attempting to share a DatabaseMetaData reference.  It fails, almost 
>>invariably, irrespective of which JDK and Oracle JDBC driver versions 
>>we try (including the 10g version).
>>
>>Funny thing is, we do have a DBA who is trying to help us analyze the 
>>problem.  As I mentioned, it's been an arduous task to replicate the 
>>problem.  However, I wish to clarify myself.  The problem with our 
>>database going down doesn't have a confirmed direct correlation to the 
>>problem we are discussing here.
>>
>>However, I am still interested in finding out if Squirrel is using the 
>>JDBC API in a way that will generate unpredictable results.  As 
>>developers use Squirrel SQL with the code completion plugin against 
>>Oracle database instances, we get SQLExceptions with messages "Protocol
> 
> 
>>Violation", "Maximum data type length exceeded", and "Connection is 
>>closed" pretty regularly as we work with the SQL editor.
>> 
>>
> 
> This particular test gives me the following:
> 
> java.sql.SQLException: Stream has already been closed
>     at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:134)
>     at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:179)
>     at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:269)
>     at oracle.jdbc.ttc7.LongTTCItem.getChars(LongTTCItem.java:189)
>     at 
> oracle.jdbc.dbaccess.DBDataSetImpl.getCharsStreamItem(DBDataSetImpl.java
> :1630)
>     at 
> oracle.jdbc.driver.OracleStatement.getCharsInternal(OracleStatement.java
> :3347)
>     at 
> oracle.jdbc.driver.OracleStatement.getStringValue(OracleStatement.java:3
> 556)
>     at 
> oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5
> 659)
>     at 
> oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5
> 622)
>     at 
> oracle.jdbc.driver.OracleResultSetImpl.getObject(OracleResultSetImpl.jav
> a:739)
>     at 
> DatabaseFailureReplicator$MetaDataTestCase.dumpResultSet(DatabaseFailure
> Replicator.java:69)
>     at 
> DatabaseFailureReplicator$MetaDataTestCase.run(DatabaseFailureReplicator
> .java:50)
>     at java.lang.Thread.run(Thread.java:534)
> 
> What I find peculiar is that the getColumns implementation (from the 
> driver I decompiled) creates a
> new PreparedStatement, executes it and returns the ResultSet, which 
> should be independent of other
> ResultSet and their associated PreparedStatements.  In other words, 
> closing one in thread A shouldn't
> have any effect on thread B's attempt to read another ResultSet.  
> Definitely requires a bit more study.
> 
> Rob
> 
> 
> CollabraSpace - Revolutionary Collaboration
> Visit us at http://www.collabraspace.com
> This message has been scanned for viruses by
> ClamAV v0.83
> 
> 
> ******* Confidentiality Notice *******
> This email, its electronic document attachments, and the contents of its website linkages may contain confidential health information.  This information is intended solely for use by the individual or entity to whom it is addressed.  If you have received this information in error, please notify the sender immediately and arrange for the prompt destruction of the material and any accompanying attachments.
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Power Architecture Resource Center: Free content, downloads, discussions,
> and more. http://solutions.newsforge.com/ibmarch.tmpl
> _______________________________________________
> Squirrel-sql-develop mailing list
> Squ...@li...
> https://lists.sourceforge.net/lists/listinfo/squirrel-sql-develop
> 

Indeed.  The stream has already been closed message is usually followed
by the other ones I mentioned in our tests.  Let me know if you need any
help in diagnosing or fixing this problem.  I would be glad to see my
favorite SQL tool become fully useable again.

Amit

-----Original Message-----
From: Rob Manning [mailto:Rob...@co...]=20
Sent: Friday, October 07, 2005 4:10 PM
To: Gollapudi, Amit C. (STL)
Cc: squ...@li...
Subject: Re: [Squirrel-sql-develop] DatabaseMetaData & Thread Safety


Gollapudi, Amit C. (STL) wrote:

>Robert:
>
>Thank you for your quick response.  I'm attaching a standalone test=20
>that we've developed to replicate the protocol violation messages we=20
>see in Squirrel sessions.  It uses pure JDBC and a set of threads=20
>attempting to share a DatabaseMetaData reference.  It fails, almost=20
>invariably, irrespective of which JDK and Oracle JDBC driver versions=20
>we try (including the 10g version).
>
>Funny thing is, we do have a DBA who is trying to help us analyze the=20
>problem.  As I mentioned, it's been an arduous task to replicate the=20
>problem.  However, I wish to clarify myself.  The problem with our=20
>database going down doesn't have a confirmed direct correlation to the=20
>problem we are discussing here.
>
>However, I am still interested in finding out if Squirrel is using the=20
>JDBC API in a way that will generate unpredictable results.  As=20
>developers use Squirrel SQL with the code completion plugin against=20
>Oracle database instances, we get SQLExceptions with messages "Protocol

>Violation", "Maximum data type length exceeded", and "Connection is=20
>closed" pretty regularly as we work with the SQL editor.
> =20
>
This particular test gives me the following:

java.sql.SQLException: Stream has already been closed
    at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:134)
    at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:179)
    at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:269)
    at oracle.jdbc.ttc7.LongTTCItem.getChars(LongTTCItem.java:189)
    at=20
oracle.jdbc.dbaccess.DBDataSetImpl.getCharsStreamItem(DBDataSetImpl.java
:1630)
    at=20
oracle.jdbc.driver.OracleStatement.getCharsInternal(OracleStatement.java
:3347)
    at=20
oracle.jdbc.driver.OracleStatement.getStringValue(OracleStatement.java:3
556)
    at=20
oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5
659)
    at=20
oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5
622)
    at=20
oracle.jdbc.driver.OracleResultSetImpl.getObject(OracleResultSetImpl.jav
a:739)
    at=20
DatabaseFailureReplicator$MetaDataTestCase.dumpResultSet(DatabaseFailure
Replicator.java:69)
    at=20
DatabaseFailureReplicator$MetaDataTestCase.run(DatabaseFailureReplicator
.java:50)
    at java.lang.Thread.run(Thread.java:534)

What I find peculiar is that the getColumns implementation (from the=20
driver I decompiled) creates a
new PreparedStatement, executes it and returns the ResultSet, which=20
should be independent of other
ResultSet and their associated PreparedStatements.  In other words,=20
closing one in thread A shouldn't
have any effect on thread B's attempt to read another ResultSet. =20
Definitely requires a bit more study.

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83


******* Confidentiality Notice *******
This email, its electronic document attachments, and the contents of its =
website linkages may contain confidential health information.  This =
information is intended solely for use by the individual or entity to =
whom it is addressed.  If you have received this information in error, =
please notify the sender immediately and arrange for the prompt =
destruction of the material and any accompanying attachments.

Gollapudi, Amit C. (STL) wrote:

>Robert:
>
>Thank you for your quick response.  I'm attaching a standalone test that
>we've developed to replicate the protocol violation messages we see in
>Squirrel sessions.  It uses pure JDBC and a set of threads attempting to
>share a DatabaseMetaData reference.  It fails, almost invariably,
>irrespective of which JDK and Oracle JDBC driver versions we try
>(including the 10g version).
>
>Funny thing is, we do have a DBA who is trying to help us analyze the
>problem.  As I mentioned, it's been an arduous task to replicate the
>problem.  However, I wish to clarify myself.  The problem with our
>database going down doesn't have a confirmed direct correlation to the
>problem we are discussing here.  
>
>However, I am still interested in finding out if Squirrel is using the
>JDBC API in a way that will generate unpredictable results.  As
>developers use Squirrel SQL with the code completion plugin against
>Oracle database instances, we get SQLExceptions with messages "Protocol
>Violation", "Maximum data type length exceeded", and "Connection is
>closed" pretty regularly as we work with the SQL editor.
>  
>
This particular test gives me the following:

java.sql.SQLException: Stream has already been closed
    at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:134)
    at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:179)
    at oracle.jdbc.dbaccess.DBError.throwSqlException(DBError.java:269)
    at oracle.jdbc.ttc7.LongTTCItem.getChars(LongTTCItem.java:189)
    at 
oracle.jdbc.dbaccess.DBDataSetImpl.getCharsStreamItem(DBDataSetImpl.java:1630)
    at 
oracle.jdbc.driver.OracleStatement.getCharsInternal(OracleStatement.java:3347)
    at 
oracle.jdbc.driver.OracleStatement.getStringValue(OracleStatement.java:3556)
    at 
oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5659)
    at 
oracle.jdbc.driver.OracleStatement.getObjectValue(OracleStatement.java:5622)
    at 
oracle.jdbc.driver.OracleResultSetImpl.getObject(OracleResultSetImpl.java:739)
    at 
DatabaseFailureReplicator$MetaDataTestCase.dumpResultSet(DatabaseFailureReplicator.java:69)
    at 
DatabaseFailureReplicator$MetaDataTestCase.run(DatabaseFailureReplicator.java:50)
    at java.lang.Thread.run(Thread.java:534)

What I find peculiar is that the getColumns implementation (from the 
driver I decompiled) creates a
new PreparedStatement, executes it and returns the ResultSet, which 
should be independent of other
ResultSet and their associated PreparedStatements.  In other words, 
closing one in thread A shouldn't
have any effect on thread B's attempt to read another ResultSet.  
Definitely requires a bit more study.

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

Robert:

Thank you for your quick response.  I'm attaching a standalone test that
we've developed to replicate the protocol violation messages we see in
Squirrel sessions.  It uses pure JDBC and a set of threads attempting to
share a DatabaseMetaData reference.  It fails, almost invariably,
irrespective of which JDK and Oracle JDBC driver versions we try
(including the 10g version).

Funny thing is, we do have a DBA who is trying to help us analyze the
problem.  As I mentioned, it's been an arduous task to replicate the
problem.  However, I wish to clarify myself.  The problem with our
database going down doesn't have a confirmed direct correlation to the
problem we are discussing here. =20

However, I am still interested in finding out if Squirrel is using the
JDBC API in a way that will generate unpredictable results.  As
developers use Squirrel SQL with the code completion plugin against
Oracle database instances, we get SQLExceptions with messages "Protocol
Violation", "Maximum data type length exceeded", and "Connection is
closed" pretty regularly as we work with the SQL editor.

Thank you for your help,
Amit


-----Original Message-----
From: Rob Manning [mailto:Rob...@co...]=20
Sent: Friday, October 07, 2005 2:39 PM
To: Gollapudi, Amit C. (STL)
Cc: squ...@li...
Subject: Re: [Squirrel-sql-develop] DatabaseMetaData & Thread Safety


Gollapudi, Amit C. (STL) wrote:

> Hi,
>
> Our development team uses Squirrel SQL to work mostly with Oracle
> databases.  Recently, however, we've been running into problems. =20
> After spending a lot of time trying to understand and reproduce the=20
> problems, we've narrowed the symptoms.=20
>
> SchemaInfo's loadColumns method gets triggered from a few places.  Of
> interest, however, is the code completion plugin.  The plugin seems to

> trigger the loadColumns on key strokes.  That is all and well.  But,=20
> loadColumns uses a class instance variable reference to=20
> DatabaseMetaData to grab column information about the table being=20
> worked on.=20
>
> Now, with Oracle's JDBC drivers, having DatabaseMetaData accessed from
> multiple threads is inherently problematic.  It causes some problems=20
> that, under the right load, take the entire database down.  As=20
> disappointing as all that is, the question is whether SchemaInfo is=20
> right in making the assumption that JDBC drivers will provide a=20
> thread-safe DatabaseMetaData object.
>
> Thoughts, suggestions?
>
I've recently decompiled the Oracle 10g JDBC driver and I notice that
oracle.jdbc.driver.OracleDatabaseMetaData has the following signature=20
for getColumns:



    public synchronized ResultSet getColumns(String s, String s1, String

s2, String s3)
        throws SQLException
  =20

Since it's synchronized, multiple threads cannot be in here at the same=20
time.  I've noticed in
the past that most public JDBC API methods in the Oracle drivers are=20
synchronized using the
instance monitor.  So, while that's a huge performance bottleneck if you

aren't careful, I don't
see that concurrent access is going to cause trouble.  I've personally=20
seen hundreds of threads
sharing access to a single pool of PreparedStatements allocated from a=20
single Connection
without any difficulty on the client.  I don't see how this could "take=20
the entire database down".
I'm not a DBA, but one might be useful in determining what statement(s)=20
are hogging the CPU/IO
on your Oracle server.

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83


******* Confidentiality Notice *******
This email, its electronic document attachments, and the contents of its =
website linkages may contain confidential health information.  This =
information is intended solely for use by the individual or entity to =
whom it is addressed.  If you have received this information in error, =
please notify the sender immediately and arrange for the prompt =
destruction of the material and any accompanying attachments.

Gollapudi, Amit C. (STL) wrote:

> Hi,
>
> Our development team uses Squirrel SQL to work mostly with Oracle 
> databases.  Recently, however, we've been running into problems.  
> After spending a lot of time trying to understand and reproduce the 
> problems, we've narrowed the symptoms. 
>
> SchemaInfo's loadColumns method gets triggered from a few places.  Of 
> interest, however, is the code completion plugin.  The plugin seems to 
> trigger the loadColumns on key strokes.  That is all and well.  But, 
> loadColumns uses a class instance variable reference to 
> DatabaseMetaData to grab column information about the table being 
> worked on. 
>
> Now, with Oracle's JDBC drivers, having DatabaseMetaData accessed from 
> multiple threads is inherently problematic.  It causes some problems 
> that, under the right load, take the entire database down.  As 
> disappointing as all that is, the question is whether SchemaInfo is 
> right in making the assumption that JDBC drivers will provide a 
> thread-safe DatabaseMetaData object.
>
> Thoughts, suggestions?
>
I've recently decompiled the Oracle 10g JDBC driver and I notice that
oracle.jdbc.driver.OracleDatabaseMetaData has the following signature 
for getColumns:



    public synchronized ResultSet getColumns(String s, String s1, String 
s2, String s3)
        throws SQLException
   

Since it's synchronized, multiple threads cannot be in here at the same 
time.  I've noticed in
the past that most public JDBC API methods in the Oracle drivers are 
synchronized using the
instance monitor.  So, while that's a huge performance bottleneck if you 
aren't careful, I don't
see that concurrent access is going to cause trouble.  I've personally 
seen hundreds of threads
sharing access to a single pool of PreparedStatements allocated from a 
single Connection
without any difficulty on the client.  I don't see how this could "take 
the entire database down".
I'm not a DBA, but one might be useful in determining what statement(s) 
are hogging the CPU/IO
on your Oracle server.

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

Hi,

Our development team uses Squirrel SQL to work mostly with Oracle
databases.  Recently, however, we've been running into problems.  After
spending a lot of time trying to understand and reproduce the problems,
we've narrowed the symptoms. =20

SchemaInfo's loadColumns method gets triggered from a few places.  Of
interest, however, is the code completion plugin.  The plugin seems to
trigger the loadColumns on key strokes.  That is all and well.  But,
loadColumns uses a class instance variable reference to DatabaseMetaData
to grab column information about the table being worked on. =20

Now, with Oracle's JDBC drivers, having DatabaseMetaData accessed from
multiple threads is inherently problematic.  It causes some problems
that, under the right load, take the entire database down.  As
disappointing as all that is, the question is whether SchemaInfo is
right in making the assumption that JDBC drivers will provide a
thread-safe DatabaseMetaData object.

Thoughts, suggestions?

Thank you,
Amit Gollapudi
Express Scripts, Inc.
******* Confidentiality Notice *******
This email, its electronic document attachments, and the contents of its =
website linkages may contain confidential health information.  This =
information is intended solely for use by the individual or entity to =
whom it is addressed.  If you have received this information in error, =
please notify the sender immediately and arrange for the prompt =
destruction of the material and any accompanying attachments.

Hi all,

I needed to add a method to the
net.sourceforge.squirrel_sql.client.preferences.IGlobalPreferencesPanel
interface.

All plugin developers who do not host their plugins in SQuirreL's CVS 
repository are asked to check if their plugins implement the interface.

Thanks

Gerd

Hello again,

This is the same row height adjustment, for the object tree.

Mitch

Hi Mitch,

your patch is in CVS.

Thanks a lot.

Gerd

Mitch Skinner schrieb:
> Hello,
> 
> I like to use large fonts, and whenever I do, the data in the dataset
> viewer gets too large for the row height.  The attached small patch adds
> a line to the DataSetViewerTablePanel.MyJTable constructor that sets the
> row height based on the font.
> 
> Regards,
> Mitch
> 
> 
> ------------------------------------------------------------------------
> 
> Index: DataSetViewerTablePanel.java
> ===================================================================
> RCS file: /cvsroot/squirrel-sql/sql12/fw/src/net/sourceforge/squirrel_sql/fw/datasetviewer/DataSetViewerTablePanel.java,v
> retrieving revision 1.27
> diff -u -r1.27 DataSetViewerTablePanel.java
> --- DataSetViewerTablePanel.java	4 Aug 2005 22:25:28 -0000	1.27
> +++ DataSetViewerTablePanel.java	24 Sep 2005 04:04:15 -0000
> @@ -171,6 +171,7 @@
>  			_typedModel = (MyTableModel) ((SortableTableModel) getModel()).getActualModel();
>  			_multiplier =
>  				getFontMetrics(getFont()).stringWidth(data) / data.length();
> +			setRowHeight(getFontMetrics(getFont()).getHeight());
>  			boolean allowUpdate = false;
>  			// we want to allow editing of read-only tables on-demand, but
>  			// it would be confusing to include the "Make Editable" option

Hello,

I like to use large fonts, and whenever I do, the data in the dataset
viewer gets too large for the row height.  The attached small patch adds
a line to the DataSetViewerTablePanel.MyJTable constructor that sets the
row height based on the font.

Regards,
Mitch

Rob Manning wrote:

>
> The file upload mechanism on SourceForge isn't working.  Until it is, 
> I won't be able
> to post the snapshots for this week on the site.  I've filed a support 
> request with SF
> which you can review (if interested) here :
>
> http://sourceforge.net/tracker/index.php?func=detail&aid=1253128&group_id=1&atid=200001 
>
>
> I'll send another email when the snapshots are available on SourceForge.

The snapshot is now available in the usual place :

http://sourceforge.net/project/showfiles.php?group_id=28383&package_id=158128

I'm not sure what my problem was with command-line anon-ftp usage was 
(it still seems broken),
but FileZilla (http://sourceforge.net/projects/filezilla/) was able to 
overcome the issue.

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

The file upload mechanism on SourceForge isn't working.  Until it is, I 
won't be able
to post the snapshots for this week on the site.  I've filed a support 
request with SF
which you can review (if interested) here :

http://sourceforge.net/tracker/index.php?func=detail&aid=1253128&group_id=1&atid=200001

I'll send another email when the snapshots are available on SourceForge.

Rob Manning


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

I have updated DAO4J plugin to v1.11 which can get the database type by JDBC 
URL. Thank you.

Zhao Liang

>From: Rob Manning <Rob...@co...>
>Reply-To: Rob...@co...
>To: ? ? <bi...@ms...>
>CC: squ...@li...
>Subject: Re: [Squirrel-sql-develop] DAO4J plugin has been released.
>Date: Tue, 02 Aug 2005 05:41:20 -0400
>
>? ? wrote:
>
>>Robert,
>>
>>Thanks your comments, I will consider it in the future releases. But it is 
>>very dificult for me to test all the scenarios of various DB and drivers.  
>>   : )
>
>I agree.  You would have to install each database and develop a schema that 
>works
>on each database.  I've done something similar in the DB Copy plugin. See 
>DialectFactory
>here:
>
>http://cvs.sourceforge.net/viewcvs.py/dbcopyplugin/src/net/sourceforge/squirrel_sql/plugins/dbcopy/dialects/DialectFactory.java?rev=1.7&view=markup
>
>Rob
>
>
>CollabraSpace - Revolutionary Collaboration
>Visit us at http://www.collabraspace.com
>This message has been scanned for viruses by
>ClamAV v0.83
>
>
>
>
>-------------------------------------------------------
>SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
>from IBM. Find simple to follow Roadmaps, straightforward articles,
>informative Webcasts and more! Get everything you need to get up to
>speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
>_______________________________________________
>Squirrel-sql-develop mailing list
>Squ...@li...
>https://lists.sourceforge.net/lists/listinfo/squirrel-sql-develop

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

? ? wrote:

> Robert,
>
> Thanks your comments, I will consider it in the future releases. But 
> it is very dificult for me to test all the scenarios of various DB and 
> drivers.    : )

I agree.  You would have to install each database and develop a schema 
that works
on each database.  I've done something similar in the DB Copy plugin. 
See DialectFactory
here:

http://cvs.sourceforge.net/viewcvs.py/dbcopyplugin/src/net/sourceforge/squirrel_sql/plugins/dbcopy/dialects/DialectFactory.java?rev=1.7&view=markup

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

Robert,

Thanks your comments, I will consider it in the future releases. But it is 
very dificult for me to test all the scenarios of various DB and drivers.    
: )

Zhao Liang

>From: Rob Manning <Rob...@co...>
>Reply-To: Rob...@co...
>To: ? ? <bi...@ms...>
>CC: squ...@li...
>Subject: Re: [Squirrel-sql-develop] DAO4J plugin has been released.
>Date: Mon, 01 Aug 2005 09:21:31 -0400
>
>? ? wrote:
>
>>Hi,
>>
>>DAO4J plugin can be used to generate Java DAO source files. You can get it 
>>from http://www.websamba.com/dao4j if you have interests. Thanks.
>>
>>Zhao Liang
>
>Hello,
>
>It's very nice!  Pretty code for being auto-generated. One enhancement you 
>might consider: I use MySQL and the
>DB Type Definition textfield was blank when I clicked on Generate DAO.  So 
>I had to browse to find the
>definition "mysql.properties".  You could "sense" the type using the 
>session driver properties (classname or vendor
>product info) and pre-populate this field.  Just a time saver, not really 
>critical.
>
>Rob
>
>
>CollabraSpace - Revolutionary Collaboration
>Visit us at http://www.collabraspace.com
>This message has been scanned for viruses by
>ClamAV v0.83
>
>
>
>
>-------------------------------------------------------
>SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
>from IBM. Find simple to follow Roadmaps, straightforward articles,
>informative Webcasts and more! Get everything you need to get up to
>speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
>_______________________________________________
>Squirrel-sql-develop mailing list
>Squ...@li...
>https://lists.sourceforge.net/lists/listinfo/squirrel-sql-develop

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

Hi,

if you agree I'd like to create a link to your plugin page here
http://squirrelsql.org/index.php?page=plugins#othersites


Gerd

? ? wrote:
> Hi,
> 
> DAO4J plugin can be used to generate Java DAO source files. You can get 
> it from http://www.websamba.com/dao4j if you have interests. Thanks.
> 
> Zhao Liang
> 
> _________________________________________________________________
> Express yourself instantly with MSN Messenger! Download today it's FREE! 
> http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
> 
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Squirrel-sql-develop mailing list
> Squ...@li...
> https://lists.sourceforge.net/lists/listinfo/squirrel-sql-develop
> 

? ? wrote:

> Hi,
>
> DAO4J plugin can be used to generate Java DAO source files. You can 
> get it from http://www.websamba.com/dao4j if you have interests. Thanks.
>
> Zhao Liang

Hello,

It's very nice!  Pretty code for being auto-generated. One enhancement 
you might consider: I use MySQL and the
DB Type Definition textfield was blank when I clicked on Generate DAO.  
So I had to browse to find the
definition "mysql.properties".  You could "sense" the type using the 
session driver properties (classname or vendor
product info) and pre-populate this field.  Just a time saver, not 
really critical.

Rob


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

Hi,

DAO4J plugin can be used to generate Java DAO source files. You can get it 
from http://www.websamba.com/dao4j if you have interests. Thanks.

Zhao Liang

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

All,

Just wanted to give you a heads up about the changes I made to build.xml 
this week.
It now uses a properties file (build.properties) to configure which 
plugins are being
built and where IzPack is installed ( instead of sections of the 
build.xml being
commented out).  Also there is a new "make_install_jar" target which 
uses the IzPack
settings in build.properties to build the installer from source.  
Additionally, I've
optimized the plugin builds so that the archives (*.tar.gz and *.zip) 
don't get
regenerated when a plugin that is configured to be rebuilt hasn't had 
any source file
changes. (Source files include doc/* lib/* src/*, etc.)  I've done a 
fair amount of
testing to ensure that *any* changes will cause the archives to get 
regenerated. 
However, if anyone notices any peculiar behavior when using the new 
build scripts,
I'd appreciate an email describing the problem.

Rob Manning


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

Rob Manning wrote:

> Brisa Francesco wrote:
>
>> hi, I'm Francesco Brisa,
>>
>> I have downloaded squirrel-sql today and tried out with linux,
>> unfortunatelly it was not starting, and I saw the problem in the 
>> squirrel-sql.sh script:
>>
>> # Squirrel home.
>> SQUIRREL_SQL_HOME='/home/robert/SQuirreL SQL Client'
>>
>> I have replaced it with:
>>
>> # Squirrel home.
>> SQUIRREL_SQL_HOME=`pwd`
>
Francesco,

I've uploaded the archive files that should address this 
problem.(squirrel-sql-2.0rc1-base.tar.gz and
squirrel-sql-2.0rc1-standard.tar.gz)

Instead of `pwd` I went with `dirname "$0"` because it makes not script 
work even when
you are not in the same directory as squirrel-sql.sh.  Give it a try 
when you have a chance, and
please let me know if it works for you.

Rob Manning


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

Brisa Francesco wrote:

> hi, I'm Francesco Brisa,
>
> I have downloaded squirrel-sql today and tried out with linux,
> unfortunatelly it was not starting, and I saw the problem in the 
> squirrel-sql.sh script:
>
> # Squirrel home.
> SQUIRREL_SQL_HOME='/home/robert/SQuirreL SQL Client'
>
> I have replaced it with:
>
> # Squirrel home.
> SQUIRREL_SQL_HOME=`pwd`

Francesco,

Thank you for pointing that out.  I'll make sure that this gets 
addressed in the archives on the download page.

Rob Manning


CollabraSpace - Revolutionary Collaboration
Visit us at http://www.collabraspace.com
This message has been scanned for viruses by
ClamAV v0.83

hi, I'm Francesco Brisa,

I have downloaded squirrel-sql today and tried out with linux,
unfortunatelly it was not starting, and I saw the problem in the 
squirrel-sql.sh script:

# Squirrel home.
SQUIRREL_SQL_HOME='/home/robert/SQuirreL SQL Client'

I have replaced it with:

# Squirrel home.
SQUIRREL_SQL_HOME=`pwd`

so now squirrel-sql can start from any directory.

Salutes

p.s.
very very good program !