Re: Change database character set
- From: Kenny Payton <k3nnyp@xxxxxxxxx>
- To: Scott Canaan <srcdco@xxxxxxx>
- Date: Fri, 24 Oct 2014 08:10:20 -0400
It sounds like you're getting off easy. As for the sys.reg$ row we had a
similar row and it was related to the emagent. We probably could have got
away with just deleting it but what we did was create a table in a user
schema with the row in it, export the user table, delete the row from both
tables and then reload the row after the csalter. In hindsight this
probably wasn't necessary but worked.
You could probably do the same with the clob data as well.
On Fri, Oct 24, 2014 at 8:04 AM, Scott Canaan <srcdco@xxxxxxx> wrote:
> Actually, I’ve been considering doing something similar to this. The 9
> rows in that table (out of about 20) that are affected were added in August
> and the person that added them still has the original documents, so they
> could be reloaded after the conversion. I believe that what is in the
> HUGECLOB in that table is just a word document that was cut and pasted into
> the field.
>
>
>
> The only row that really concerns me is the one in SYS.REG$.
>
>
>
> Scott Canaan ’88 (srcdco@xxxxxxx)
>
> (585) 475-7886 – work (585) 339-8659 – cell
>
> “Life is like a sewer, what you get out of it depends on what you put into
> it.” – Tom Lehrer
>
>
>
> *From:* Chris Taylor [mailto:christopherdtaylor1994@xxxxxxxxx]
> *Sent:* Friday, October 24, 2014 8:01 AM
> *To:* Niall Litchfield
> *Cc:* Scott Canaan; oracle.unknowns@xxxxxxxxx; oracle-l@xxxxxxxxxxxxx
> *Subject:* Re: Change database character set
>
>
>
> Based on Niall's analysis, another option could be to export the
> SCRIPT_GROUP table, truncate it, do the conversion and reload it (depending
> on the size).
>
>
>
> Chris
>
>
>
>
>
> On Fri, Oct 24, 2014 at 5:02 AM, Niall Litchfield <
> niall.litchfield@xxxxxxxxx> wrote:
>
> Scott
>
>
>
> I'm pretty sure you'll find the session_key row is transient and due to an
> emagent (either grid/cloud control or dbconsole) - there's a note on csscan
> dictionary errors that will tell you. That leaves you with 9 documents
> (they look like word docs/emails). Can you grab those from the app and
> reenter the data post conversion?
>
>
>
Other related posts:
