You are viewing a plain text version of this content. The canonical link for it is here.
Posted to ojb-user@db.apache.org by Edson Carlos Ericksson Richter <ed...@mgrinformatica.com.br> on 2004/11/25 13:08:59 UTC
Connection broken
Hi, guys.
I'm deploying a really big app written using OJB (today is about 360000
loc), that uses MaxDB as database server.
I've a test connection query (select 1 from DBA.DUAL), and it works.
But after some time of work (about 4 hours of continuous use), OJB loose
one connection, and never get it working again. May be this a known bug
for OJB 1.0.1? Should I upgrade DBCP (I'm using it)? I've checked past
threads, but just only one appear be related... And tests exchanging
DBCP to ConnectionPool don't solved. Using no connection pool appear be
solved, but our app is high query/transaction intensive, and opening a
connection is an issue for us (we use a pool of 10 connections per user
here, and we have ~40 users)).
Thanks for any help,
Edson Richter
---------------------------------------------------------------------
To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
For additional commands, e-mail: ojb-user-help@db.apache.org
Re: Connection broken
Posted by Thomas Paradies <pa...@transit-online.de>.
Hi,
this is not a problem with OJB or connection pooling. I think this is caused by
a timeout setting in MaxDB. If you are working with SQL Studio after some time
you'll notice that you have to perform actions twice to take effect because you
loose your session/connection. However that may be the following provisional url
setting did it for us:
-> jdbc:sapdb://host/DB?timeout=0
But try it with caution,
Thomas
Edson Carlos Ericksson Richter wrote:
> Hi, guys.
>
> I'm deploying a really big app written using OJB (today is about 360000
> loc), that uses MaxDB as database server.
> I've a test connection query (select 1 from DBA.DUAL), and it works.
>
> But after some time of work (about 4 hours of continuous use), OJB loose
> one connection, and never get it working again. May be this a known bug
> for OJB 1.0.1? Should I upgrade DBCP (I'm using it)? I've checked past
> threads, but just only one appear be related... And tests exchanging
> DBCP to ConnectionPool don't solved. Using no connection pool appear be
> solved, but our app is high query/transaction intensive, and opening a
> connection is an issue for us (we use a pool of 10 connections per user
> here, and we have ~40 users)).
>
> Thanks for any help,
>
> Edson Richter
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
> For additional commands, e-mail: ojb-user-help@db.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
For additional commands, e-mail: ojb-user-help@db.apache.org
Re: Connection broken
Posted by Edson Carlos Ericksson Richter <ed...@mgrinformatica.com.br>.
I'll do checks and answer in a few hours (it take few hours to I get the
error again - it's not determinate).
Thank you for directions.
Richter
Armin Waibel escreveu:
> Hi Edson,
>
> Edson Carlos Ericksson Richter wrote:
>
>> Hi, guys.
>>
>> I'm deploying a really big app written using OJB (today is about
>> 360000 loc), that uses MaxDB as database server.
>> I've a test connection query (select 1 from DBA.DUAL), and it works.
>>
>> But after some time of work (about 4 hours of continuous use), OJB
>> loose one connection, and never get it working again. May be this a
>> known bug for OJB 1.0.1? Should I upgrade DBCP (I'm using it)?
>
>
> It's not a known bug. Other user posted about an issue when using lazy
> collection references, but I never notice an issue about "lost
> connections". How does OJB loose the connection? Will it be removed
> from pool without closing or never returned to pool and timed out? Did
> you get any error message, stack trace?
>
>
>> I've checked past threads, but just only one appear be related... And
>> tests exchanging DBCP to ConnectionPool don't solved.
>
>
> Both implementations ConnectionFactoryDBCPImpl and
> ConnectionFactoryPooledImpl rely on commons-pool, so they should
> behave in same way.
>
>
>> Using no connection pool appear be solved, but our app is high
>> query/transaction intensive, and opening a connection is an issue for
>> us (we use a pool of 10 connections per user here, and we have ~40
>> users)).
>
>
> This indicate that OJB does a good job and close all used connections,
> but maybe the issue only disappear because of reduced performance.
> Do you use the 'maxIdle' property in ConnectionFactoryPooled/DBCPImpl?
>
>
> regards,
> Armin
>
>>
>> Thanks for any help,
>>
>> Edson Richter
>>
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
>> For additional commands, e-mail: ojb-user-help@db.apache.org
>>
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
> For additional commands, e-mail: ojb-user-help@db.apache.org
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
For additional commands, e-mail: ojb-user-help@db.apache.org
Re: Connection broken
Posted by Armin Waibel <ar...@apache.org>.
Hi Edson,
Edson Carlos Ericksson Richter wrote:
> Hi, guys.
>
> I'm deploying a really big app written using OJB (today is about 360000
> loc), that uses MaxDB as database server.
> I've a test connection query (select 1 from DBA.DUAL), and it works.
>
> But after some time of work (about 4 hours of continuous use), OJB loose
> one connection, and never get it working again. May be this a known bug
> for OJB 1.0.1? Should I upgrade DBCP (I'm using it)?
It's not a known bug. Other user posted about an issue when using lazy
collection references, but I never notice an issue about "lost
connections". How does OJB loose the connection? Will it be removed from
pool without closing or never returned to pool and timed out? Did you
get any error message, stack trace?
> I've checked past
> threads, but just only one appear be related... And tests exchanging
> DBCP to ConnectionPool don't solved.
Both implementations ConnectionFactoryDBCPImpl and
ConnectionFactoryPooledImpl rely on commons-pool, so they should behave
in same way.
> Using no connection pool appear be
> solved, but our app is high query/transaction intensive, and opening a
> connection is an issue for us (we use a pool of 10 connections per user
> here, and we have ~40 users)).
This indicate that OJB does a good job and close all used connections,
but maybe the issue only disappear because of reduced performance.
Do you use the 'maxIdle' property in ConnectionFactoryPooled/DBCPImpl?
regards,
Armin
>
> Thanks for any help,
>
> Edson Richter
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
> For additional commands, e-mail: ojb-user-help@db.apache.org
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: ojb-user-unsubscribe@db.apache.org
For additional commands, e-mail: ojb-user-help@db.apache.org