On Jan 31, 2007, at 6:49 PM, Kris Jurka wrote:
>
>
> On Wed, 31 Jan 2007, Jeffrey Cox wrote:
>
>> Yea sure... I will look into the COLUM_TYPE issue and write a test
>> case for column names. On that note, when I ran the test suit all
>> the jdbc2 test would run fine, however, most of the others failed
>> due to "to many connections". I of course tried to up my max
>> connections beyond 20, but looks like will need to up shared
>> memory maximums. (which might be interesting given I am on os X.
>> Just never done that since switching). So before I go about
>> figuring all that out, are failures expected? and if not, any
>> thoughts on the number of connections I should configure for?
>>
>
> No failures are expected, but it's not surprising for things to go
> bad after completing the jdbc2 tests because that's when it starts
> testing connection pooling stuff that will open a fair number of
> connections. I don't think it uses more than 20 simultaneous
> connections so it's likely that there are just some that aren't
> closed or are in the process of closing as new ones are opened.
>
> If you can't get these tests straightened out I wouldn't worry
> about it because the changes you are making shouldn't affect any
> tests beyond the jdbc2 ones.
>
Yes I assumed as much.
I took a look at the the code to night regarding COLUMN_TYPE, and it
doesn't look to bad. Given other project work this week, it will
probably be the weekend before I churn it out.
Jeff
> Kris Jurka