Обсуждение: Problem in restoring data

Hi,

 I face the following problem in transferring data from
pgsql 7.3.4 to pgsql 7.4 RC1 . I am piping the output of
pg_dumpall from 7.3 to 7.4 running on different port.

The problem is there is a gist index on txtidx type on a
non-public schema  and when search_path does not include
public  the index cannot be created.

I solution i have is to drop such  indexes temporarily before
dumping. Is that the only way ?

PS: i do not want to edit the (splitted) dump file , my database
is ~ 7 GB


regds
mallah.



...........
...........

CREATE INDEX intranet_byline_ed ON intranet_byline USING btree (edition);
CREATE INDEX
CREATE INDEX iid_listing_userid ON iid_listing USING btree (userid);
CREATE INDEX
CREATE INDEX intranet_company_category_name ON intranet_company USING
btree (category_name);
CREATE INDEX
*SET search_path = archives, pg_catalog;*
SET
CREATE INDEX site_search_gist ON site_search USING gist (keywordidx);
*ERROR:  data type public.txtidx has no default operator class for
access method "gist"*
HINT:  You must specify an operator class for the index or define a
default operator class for the data type.
SET search_path = data_bank, pg_catalog;
SET
CREATE INDEX profiles_pincode ON profiles USING btree (pincode);
CREATE INDEX
CREATE INDEX profiles_city ON profiles USING btree (city);
CREATE INDEX
CREATE INDEX profiles_branch ON profiles USING btree (branch_id);
CREATE INDEX
CREATE INDEX profiles_area_code ON profiles USING btree (area_code);
CREATE INDEX
...........
...........

On Mon, Nov 10, 2003 at 01:36:20AM +0530, Rajesh Kumar Mallah wrote:

> PS: i do not want to edit the (splitted) dump file , my database
> is ~ 7 GB

You can "edit" the dump file by means of
perl -pi -e 's/^CREATE INDEX.*gist.*//' dump-file

or something like that ...
(this isn't a solution to your problem, but may serve as workaround)

--
Alvaro Herrera (<alvherre[a]dcc.uchile.cl>)
"Lo esencial es invisible para los ojos" (A. de Saint Exúpery)

On Mon, Nov 10, 2003 at 01:36:20AM +0530, Rajesh Kumar Mallah wrote:
 

PS: i do not want to edit the (splitted) dump file , my database
is ~ 7 GB   

You can "edit" the dump file by means of
perl -pi -e 's/^CREATE INDEX.*gist.*//' dump-file

or something like that ...
(this isn't a solution to your problem, but may serve as workaround)
 

Rajesh Kumar Mallah <mallah@trade-india.com> writes:
>  I face the following problem in transferring data from
> pgsql 7.3.4 to pgsql 7.4 RC1 . I am piping the output of
> pg_dumpall from 7.3 to 7.4 running on different port.
> The problem is there is a gist index on txtidx type on a
> non-public schema  and when search_path does not include
> public  the index cannot be created.

There is a post-7.3.4 bug fix in the 7.3 branch for this mistake:

2003-10-02 18:25  tgl

    * src/backend/utils/adt/ruleutils.c (REL7_3_STABLE): When dumping
    CREATE INDEX, must show opclass name if the opclass isn't in the
    schema search path.  Otherwise pg_dump doesn't correctly dump
    scenarios where a custom opclass is created in 'public' and then
    used by indexes in other schemas.

Since the bug is in the backend and not pg_dump, you can't escape it by
using the 7.4 version of pg_dump against the 7.3 server.  You could
recompile the server using the 7.3-branch-head version of ruleutils.c,
though.

            regards, tom lane

Tom Lane wrote:

>Rajesh Kumar Mallah <mallah@trade-india.com> writes:
>
>
>> I face the following problem in transferring data from
>>pgsql 7.3.4 to pgsql 7.4 RC1 . I am piping the output of
>>pg_dumpall from 7.3 to 7.4 running on different port.
>>The problem is there is a gist index on txtidx type on a
>>non-public schema  and when search_path does not include
>>public  the index cannot be created.
>>
>>
>
>There is a post-7.3.4 bug fix in the 7.3 branch for this mistake:
>
>2003-10-02 18:25  tgl
>
>    * src/backend/utils/adt/ruleutils.c (REL7_3_STABLE): When dumping
>    CREATE INDEX, must show opclass name if the opclass isn't in the
>    schema search path.  Otherwise pg_dump doesn't correctly dump
>    scenarios where a custom opclass is created in 'public' and then
>    used by indexes in other schemas.
>
>Since the bug is in the backend and not pg_dump, you can't escape it by
>using the 7.4 version of pg_dump against the 7.3 server.
>

Ok,I read somewhere its always better to use more recent pg_dump while
migrating.

>  You could
>recompile the server using the 7.3-branch-head version of ruleutils.c,
>

Thanks for the explanation , Shall do that  please tell me how to fetch
ruleutils.c
from 7.3-branch-head (do not know the cvs commands   O:-) ) , thanks again.

regds
mallah.

>though.
>
>            regards, tom lane
>
>---------------------------(end of broadcast)---------------------------
>TIP 6: Have you searched our list archives?
>
>               http://archives.postgresql.org
>
>

Rajesh Kumar Mallah <mallah@trade-india.com> writes:
> Thanks for the explanation , Shall do that  please tell me how to fetch
> ruleutils.c
> from 7.3-branch-head (do not know the cvs commands   O:-) ) , thanks again.

The easiest way to get it is from the cvsweb interface:
http://developer.postgresql.org/cvsweb.cgi/pgsql-server/src/backend/utils/adt/ruleutils.c
Shift-click on the "download" link for the revision you want (1.124.2.2)
and you got it.

            regards, tom lane

Rajesh Kumar Mallah <mallah@trade-india.com> writes: 

Thanks for the explanation , Shall do that  please tell me how to fetch 
ruleutils.c
from 7.3-branch-head (do not know the cvs commands   O:-) ) , thanks again.   

The easiest way to get it is from the cvsweb interface:
http://developer.postgresql.org/cvsweb.cgi/pgsql-server/src/backend/utils/adt/ruleutils.c
Shift-click on the "download" link for the revision you want (1.124.2.2)
and you got it.
		regards, tom lane