Re: Weird listbox issue (Gupta(SqlWindows), Oracle)

From: Shakespeare <>
Date: Fri, 05 Dec 2008 18:17:41 +0100
Message-ID: <4939623f$0$188$>

MVk - Monika V. kycka schreef:
> On 5 Gruo, 00:20, joel garry <> wrote:

>> On Dec 4, 11:52 am, "MVk - Monika V. kycka" <>
>> wrote:
>> You can believe it when I say "I hear you."  :-)
>> I can guess there will never be less work, but I can also guess there
>> will be a lot more work due to issues like this.  If the client
>> doesn't have a DBA available that understands these issues they are
>> asking for trouble.
>> jg
>> --
>> is bogus.
>> “Because there is no job guarantees for IT people, for the last six
>> months brides' families have not been accepting grooms from this
>> background.” - Jagadeesh Angadi, a matchmaker in Bangalore.

> Ha! I know the reason, I was right from the beginning! :)
> I noticed a bit different application session representation in PlSql
> Developer. Application selects where shown not with variables like :1,
> but like :"SYS_B_0". It looked a bit weird for me, so I tried seaching
> information about it. I found it was related to CURSOR_SHARING
> parameter. Indeed, it was different for the databases - test db had
> 'EXACT', production - 'FORCE'. I tried changing it on one of our local
> test dbs, but it didn't work out - application still was showing data
> correctly. So I started looking for other possible reasons, wrote
> here.
> And today another weird activity occured for clients production db -
> not only filters not filling, but no one could connect to db with
> application at all. Then the clien says, 'Oh... you know, some long
> time ago there was something like that. We had to change one db
> parameter. We'll try changing it'. So they did change it and hurray -
> everything is working just fine! :) Indeed, it was the fault of
> CURSOR_SHARING. It had to be changed from 'FORCE' to 'EXACT'.
> I'm releaved now that I'm definite about the reason and that actualy I
> was looking rigth at this from the beginning. Pitty, I did something
> wrong and missed it, but still... I'm so happy - problem solved! :)
> MVk - Monika V. kycka

Thanks for the feedback. Another proof of faulty results with cursor sharing = forced!

Shakespeare Received on Fri Dec 05 2008 - 11:17:41 CST

Original text of this message