Mailing-List: contact derby-dev-help@db.apache.org; run by ezmlm
Precedence: bulk
Reply-To: <derby-dev@db.apache.org>
Received-SPF: pass (nike.apache.org: local policy)
Date: Thu, 06 Sep 2007 02:07:00 +0200
From: Dag.Wanvik@Sun.COM (Dag H. Wanvik)
Subject: Re: [jira] Commented: (DERBY-3021) Replication: Add a
 ReplicationSlave
 controller that will manage replication on the slave side
In-reply-to: <14263519.1188985113106.JavaMail.root@brutus>
Sender: Dag.Wanvik@Sun.COM
To: derby-dev@db.apache.org
Message-id: <x1oxhcm8mzbv.fsf@sun.com>
MIME-version: 1.0
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: QUOTED-PRINTABLE
References: <14263519.1188985113106.JavaMail.root@brutus>
User-Agent: Gnus/5.1008 (Gnus v5.10.8) Emacs/22.0.50 (usg-unix-v)

"J=F8rgen L=F8land (JIRA)" <jira@apache.org> writes:

> Any thoughts on whether the most appropriate error code for this wo=
uld be:
>
> * XRE02.C since it is a replication error message with session seve=
rity, or
> * 0800x (e.g., 08006 - connection failure) since this is a connecti=
on error?=20


08006 seems to be a network client error, wouldn't a new 08004 (the
server is there but rejects the connection for some reason...) be mor=
e
in the spirit of existing allocations? But I am OK with XRE02.C, too.

I looked at some other non-"08xxx" which have session severity level
in Derby, but am not sure they give much guidance. Maybe we should
prefer 08004 when reasonble, since this has standard SQL semantics
(not implementation defined). In JDBC 4 this SQLSTATE class has a
defined mapping to SQL(Non)TransientConnectionExceptions, too.

I must admit I feel a bit out on a limb, here, so it someone has any
general guidelines for how to allocate new error codes in Derby, I,
for one, would appreciate it.

Dag