What I was suggesting is to do a "kill -QUIT <pid>" on the JVM (Ctrl-
Break on Windows) to see what the threads are doing during that period.
http://java.sun.com/developer/technicalArticles/Programming/Stacktrace/
Andrus
On Jan 30, 2006, at 9:51 PM, Bryan Lewis wrote:
> A nonSeletingQuery didn't change it. I made sure I was using the
> 1.2M10
> jar. Mysterious, I'll try to put together a small test case
> tomorrow.
>
> There is no stack dump worth seeing. The log shows "[QueryLogger] ===
> updated 20 rows" forever, until I kill the server.
>
>
> Andrus Adamchik wrote:
>
>> Hmm... I can't reproduce it with M10 or the current CVS code. If you
>> can provide a test case (or maybe a stack dump?) it would be great.
>>
>> It shouldn't really matter (I tested it both ways), but does it work
>> if you replace "dc.performQuery(query)" with
>> "dc.performNonSelectingQuery(query)"?
>>
>> Andrus
>>
>>
>>
>> On Jan 30, 2006, at 7:32 PM, Bryan Lewis wrote:
>>
>>> I have some code that worked well in earlier versions of Cayenne
>>> (1.2M4
>>> or before) but now it goes into an infinite loop.
>>>
>>> String sql = "delete from Search_Data where UserName = 'joe'";
>>> SQLTemplate query = new SQLTemplate(Company.class, sql);
>>> dc.performQuery(query);
>>>
>>> The query is deleting all the rows from the user's previous
>>> search. In
>>> this case there are 20 rows.
>>>
>>> The code enters the performQuery() but never comes out. The log
>>> shows
>>> an infinite list of this message:
>>>
>>> [QueryLogger] === updated 20 rows.
>>>
>>>
>>>
>>
>
>
This archive was generated by hypermail 2.0.0 : Mon Jan 30 2006 - 22:26:59 EST