On 27/04/11 15:33, Ulrich Weiss wrote:
Hello!
Yes, the problem exists also with DBD::Oracle 1.27.
With my test machine, I have checked following environments:
Failed:
11R2, DBD 1.28, DBI-1.613-1.fc14
11R2, DBD 1.27, DBI-1.613-1.fc14
11R2, DBD 1.22, DBI-1.613-1.fc14
11R2, DBD 1.22, DBI-1.607 (RHEL 5.3, 64bit)
Ok:
11R1, DBD 1.28, DBI-1.613-1.fc14
11R1, DBD 1.22, DBI-1.613-1.fc14
11R1, DBD 1.22, DBI-1.607 (RHEL 5.3, 64bit)
Am Mittwoch, 27. April 2011, 15:53:51 schrieben Sie:
There was some changes on SIG handling in 1.28 (you will have to ask Martin
about the details it was his fix)
I read your Oracle post. Just to fully remove DBI and DBD::Oracle from the
picture can you try the same code/oracle 11g client but this time use
DBD::Oracle 1.27
That would but it all in Oracle's court then
Cheers
John Scoles
From: ulrich.we...@gmx.at
To: dbi-users@perl.org
Subject: Oracle 11R2, perl-DBD: eating signal CTRL-C
Date: Wed, 27 Apr 2011 15:42:40 +0200
Hello!
I'm new in this list, but I have found a problem, perhaps a bug in perl
DBD. Please can you look at following link:
http://forums.oracle.com/forums/thread.jspa?messageID=9521449
I thought of a problem in the oracle-lib, but it may be (also) in DBD.
Can someone please tell me anything about this behaviour?
greetings
Ulrich Weiss
This is an old issue but the signals involved change depending on the version
of the oracle client. Sometimes, SIGINT, sometimes SIGCHLD - others have
reported other signals (search this list for +signal +oracle and you'll see
many). Often it depends on how you connect.
We have the following old knowledge base article on our site -
http://www.easysoft.com/support/kb/kb00967.html
You can use ora_connect_with_default_signals to save the signal disposition but
I'm always wary of this this - the client libs must be using the signal for
something.
Martin
--
Martin J. Evans
Easysoft Limited
http://www.easysoft.com
