ibatis-user-java mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From cmose <cjm...@gmail.com>
Subject Re: Deadlock found when trying to get lock; try restarting transaction
Date Fri, 02 Feb 2007 17:36:51 GMT

clinton, thanks I think that might be the case... I went ahead and added some
exception trapping to handle the deadlocks and the more recent occurences of
Lock Timeout messages such that the code has a max # of retries and sleeps
in between those. So far that seems to be working.

Any idea on the Lock Timeout error? That is a new one to me and just started
showing up today, in addition to the deadlock issue...

thanks again for the help.


Clinton Begin wrote:
> 
> yes...that could be a problem.  When you ctrl-c, it may not close the
> connections properly.  Not sure how MySQL handles that.
> 
> 
> 
> On 2/2/07, cmose <cjmose@gmail.com> wrote:
>>
>>
>> Clinton, thanks for the advice, I did remove those settings, however,
>> that
>> did not appear to have any noticeable effect on the problem.
>>
>> I'm not really sure how I could be encountering a situation where a
>> transaction is not ended properly - transactional code only shows up 1
>> place
>> in my code and that is wrapped with a finally{dao.endTransaction()}....I
>> have noticed that the problem tends to occur more frequently when I
>> control-c the application and then restart it, I'm not sure how that
>> could
>> be relevant though....
>>
>>
>>
>> Clinton Begin wrote:
>> >
>> > Hi there,
>> >
>> > I'd recommend getting rid of those settings altogether.  While they
>> very
>> > well could be the cause, they are not the solution.  It's more likely
>> that
>> > you have some code that's not ending a transaction correctly.
>> >
>> > The defaults are very high, but the relative values are more important
>> > than
>> > the size of the values.
>> >
>> > In the future I may suggest we remove these settings altogether.  They
>> > were
>> > a performance benefit for some extremely constrained architectures,
>> like
>> > the
>> > onese in the PetStore performance testing....it's all a bit academic
>> and
>> > less than practical.
>> >
>> > Clinton
>> >
>> > On 2/1/07, cmose <cjmose@gmail.com> wrote:
>> >>
>> >>
>> >> Hello all,
>> >> I've got fairly staright forward application in which there ~ 3 -15
>> >> threads
>> >> running that hit a database. The threads can potentially try to update
>> >> the
>> >> same table. I'm just getting into the test stage and have hit a snag
>> with
>> >> ibatis and transactions. I'll randomly receive
>> >> MySQLTransactionRollbackExceptions at different points thoughout the
>> >> application and I'm not really sure where to look for the
>> culprit...Any
>> >> advice is very welcome!
>> >>
>> >> here is the settings tag for my sqlmapconfig file
>> >> <settings catcheModelsEnabled="false" enchancementEnabled="true"
>> >> maxSessions="80" maxTransactions="40" maxRequests="128"
>> >> useStatementNamespaces="true" />
>> >>
>> >> The numbers are a bit wonky as I've been fiddling with them to see if
>> >> they
>> >> were the cause. If anyone could point me in the right direction of
>> what
>> I
>> >> should be looking I would appreciate it - if more information is
>> needed,
>> >> please let me know.
>> >> Thanks!
>> >> --
>> >> View this message in context:
>> >>
>> http://www.nabble.com/Deadlock-found-when-trying-to-get-lock--try-restarting-transaction-tf3157415.html#a8756469
>> >> Sent from the iBATIS - User - Java mailing list archive at Nabble.com.
>> >>
>> >>
>> >
>> >
>>
>> --
>> View this message in context:
>> http://www.nabble.com/Deadlock-found-when-trying-to-get-lock--try-restarting-transaction-tf3157415.html#a8769747
>> Sent from the iBATIS - User - Java mailing list archive at Nabble.com.
>>
>>
> 
> 

-- 
View this message in context: http://www.nabble.com/Deadlock-found-when-trying-to-get-lock--try-restarting-transaction-tf3157415.html#a8771629
Sent from the iBATIS - User - Java mailing list archive at Nabble.com.


Mime
View raw message