headers
fabrice.bourdel | 1 Sep 2004 03:18
Picon
Favicon

pb with recursive call to procedures ?

hello,

I'm having a stange problem into a process (stored procedure).

The purpose of this process is to generate a tree representation
from a hierarchical representation.

The process is recursive, and it seems i am loosing the position
of the cursor by returning back from the recursive call (after
one branch has been processed)
In the slight exemple below, just A, B and C would be processed.

I saw that unamed result tables have different "temporary" names
and that's the reason why i don't understand that the process
wouldn't be ok. Also with named results (with a declare xx cursor)
the problem wouldn't change.

When a procedure returns a cursor, the name can be set. But in
my case, would it be possible ?

bests regards.

***************************************************
about the datas
***************************************************

The hierarchical representation is :
 - id  : internal number
 - lib : text
 - pere: father id
(Continue reading)

Permalink | Reply | 0 comments |
headers
Tomas Simecek | 1 Sep 2004 09:20
Picon

Order by

Hi,

I have table

create table t1
(
i1 integer,
i2 integer
)

Following select is OK.
select i1 from t1 order by i2

But this one is problematic:
select distinct i1 from t1 order by i2

Gives me
General error;-8013 POS(37) Order column must be output column.

Why this error ? Bug ?

Thanks,
    Tom

--

-- 
MaxDB Discussion Mailing List
For list archives: http://lists.mysql.com/maxdb
To unsubscribe:    http://lists.mysql.com/maxdb?unsub=gcdm-maxdb <at> m.gmane.org
(Continue reading)

Permalink | Reply | 0 comments |
headers
Zabach, Elke | 1 Sep 2004 09:07
Picon
Favicon

AW: Order by

Tomas Simecek wrote:
> 
> Hi,
> 
> I have table
> 
> create table t1
> (
> i1 integer,
> i2 integer
> )
> 
> Following select is OK.
> select i1 from t1 order by i2
> 
> But this one is problematic:
> select distinct i1 from t1 order by i2
> 
> Gives me
> General error;-8013 POS(37) Order column must be output column.
> 
> Why this error ? Bug ?

No bug, but 'feature'. It has to do with the implementation of ORDER BY and DISTINCT that the order by columns
have to be part of the output if DISTINCT is specified.

Elke
SAP Labs Berlin

>
(Continue reading)

Permalink | Reply | 501 comments |
headers
Xavier CAMIER | 1 Sep 2004 09:08
Picon

Re: AW: AW: Cascade Triggering

Anhaus, Thomas a écrit :

>Hi Xavier,
>I'm interested in the sql code of the failing update. Since your
>trigger terminates with a STOP statement in case of an error,
>the update statement should return with that trigger error code.
>
>Kind Regards,
>Thomas
>
>  
>
>>-----Ursprüngliche Nachricht-----
>>Von: Xavier CAMIER [mailto:xavier.camier <at> batinor.fr] 
>>Gesendet: Dienstag, 31. August 2004 17:37
>>An: Anhaus, Thomas; maxdb <at> lists.mysql.com
>>Betreff: Re: AW: Cascade Triggering
>>
>>
>>Anhaus, Thomas a écrit :
>>
>>    
>>
>>>Xavier CAMIER wrote : 
>>>
>>> 
>>>
>>>      
>>>
>>>>Good morning,
(Continue reading)

Permalink | Reply | 0 comments |
headers
Arne Gehlhaar | 1 Sep 2004 09:43
Favicon

Re: AW: Order by

Zabach, Elke wrote:
> Tomas Simecek wrote:
> 
>>Hi,
>>
>>I have table
>>
>>create table t1
>>(
>>i1 integer,
>>i2 integer
>>)
>>
>>Following select is OK.
>>select i1 from t1 order by i2
>>
>>But this one is problematic:
>>select distinct i1 from t1 order by i2
>>
>>Gives me
>>General error;-8013 POS(37) Order column must be output column.
>>
>>Why this error ? Bug ?
> 
> 
> No bug, but 'feature'. It has to do with the implementation of ORDER BY and DISTINCT that the order by
columns have to be part of the output if DISTINCT is specified.

Seems to me a very reasonable limitation. How would you like to sort the 
following otherwise:
(Continue reading)

Permalink | Reply | 501 comments |
headers
Xavier CAMIER | 1 Sep 2004 14:55
Picon

Re: AW: AW: AW: Cascade Triggering

Anhaus, Thomas a écrit :

>Hi Xavier,
>error -4000 gave me a hint, but no explanation. Error -4000 means 'Unknown result table'
>and I think that one of the statements
>
>CLOSE CURSCUMULPARTIEL; 
>or
>CLOSE CURSCUMULFINAL;
>
>produces the error. But I have no idea, why these cursors should not exist
>any more.
>I fear only a vtrace give us more information (For vtrace please have a look to
>http://sapdb.2scale.net/moin.cgi/VTrace).
>
>Alternatively could try to ignore error -4000 inside your db-procedure.
>
>Best Regards,
>Thomas
>
>
>  
>
>>-----Ursprüngliche Nachricht-----
>>Von: Xavier CAMIER [mailto:xavier.camier <at> batinor.fr] 
>>Gesendet: Mittwoch, 1. September 2004 09:08
>>An: Anhaus, Thomas; maxdb <at> lists.mysql.com
>>Betreff: Re: AW: AW: Cascade Triggering
>>
>>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Kevin.Wilson | 1 Sep 2004 15:49

strange connection reports in knldiag

What kind of state would a servlet app using the sapdb jdbc driver have to
be in to cause this type of log entry? Also, I have no apps running locally
on the db server so is the local connection shown from a servertask thread? 

2004-08-31 19:34:52 16217     11007 COMMUNIC wait for connection T41
2004-08-31 19:34:52 16217     11561 COMMUNIC Connected  T41 bobo.mycom.com 0
2004-08-31 20:04:59 16217     11560 COMMUNIC Releasing  T38
2004-08-31 20:04:59 16217     11560 COMMUNIC Releasing  T39
2004-08-31 20:04:59 16217     11007 COMMUNIC wait for connection T38
2004-08-31 20:04:59 16217     11007 COMMUNIC wait for connection T39
2004-08-31 20:04:59 16217     11560 COMMUNIC Releasing  T40
2004-08-31 20:04:59 16217     11007 COMMUNIC wait for connection T40
2004-08-31 20:04:59 16217     11560 COMMUNIC Releasing  T36
2004-08-31 20:04:59 16217     11007 COMMUNIC wait for connection T36
2004-08-31 20:04:59 16217     11560 COMMUNIC Releasing  T37
2004-08-31 20:04:59 16217     11007 COMMUNIC wait for connection T37
2004-08-31 20:04:59 16217     11509 COMMUNIC Connection broken by appl state
10 T41
2004-08-31 20:04:59 16217 WNG 11824 COMMUNIC Releasing  T41 connection
aborted
2004-08-31 20:04:59 16217     11007 COMMUNIC wait for connection T41
2004-08-31 20:40:31 16217     11561 COMMUNIC Connected  T36 momo.mycom.com 0
2004-08-31 20:40:31 16217     11560 COMMUNIC Releasing  T36
2004-08-31 20:40:31 16217     11007 COMMUNIC wait for connection T36
2004-08-31 20:40:31 16217     11561 COMMUNIC Connected  T36 momo.mycom.com 0
2004-08-31 21:04:46 16217     11561 COMMUNIC Connected  T37 bobo.mycom.com 0
2004-08-31 21:04:46 16217     11561 COMMUNIC Connected  T38 bobo.mycom.com 0
2004-08-31 21:04:46 16217     11561 COMMUNIC Connected  T39 bobo.mycom.com 0
2004-08-31 21:04:46 16217     11561 COMMUNIC Connected  T40 bobo.mycom.com 0
2004-08-31 21:04:46 16217     11561 COMMUNIC Connected  T41 bobo.mycom.com 0
(Continue reading)

Permalink | Reply | 0 comments |
headers
Ruth, Brice | 1 Sep 2004 17:39

MAPCHAR + UNICODE

Just a heads up for the developers/documentation developers. I did some 
googling on MaxDB + MAPCHAR in UNICODE databases and found many 
references on the mailing list (from Elke) indicating that this is 
basically not supported. I wanted to point out a documentation link that 
disagrees:

http://www.sapdb.org/7.4/htmhelp/be/a51a26c6cf11d5993e00508b6b8b11/frameset.htm

Or, at least it seems to disagree to me. I did verify, however, that in 
the DB Manager GUI, I cannot create a UNICODE MAPCHAR definition 
(unfortunately).

I looked into this because we have an application that is displaying 
German cities and Öttingen is being displayed at the very end of the 
list, because the column is (and needs to be) UNICODE.

It would certainly be nice to be able to address this efficiently at the 
DB level, w/o needing to add sort logic to the application.

Respectfully,
Brice Ruth

-- 
Brice Ruth, Sr. IT Analyst
Fiskars Brands Inc
http://www.fiskarsbrands.com/

--

-- 
MaxDB Discussion Mailing List
For list archives: http://lists.mysql.com/maxdb
(Continue reading)

Permalink | Reply | 0 comments |
headers
Schroeder, Alexander | 1 Sep 2004 17:57
Picon
Favicon

RE: MAPCHAR + UNICODE

Hello Ruth, 

> Just a heads up for the developers/documentation developers. 
> I did some 
> googling on MaxDB + MAPCHAR in UNICODE databases and found many 
> references on the mailing list (from Elke) indicating that this is 
> basically not supported. I wanted to point out a 
> documentation link that 
> disagrees:
> 
> http://www.sapdb.org/7.4/htmhelp/be/a51a26c6cf11d5993e00508b6b
> 8b11/frameset.htm

Thank you for pointing this out, it seems to be wrong ...

> Or, at least it seems to disagree to me. I did verify, 
> however, that in 
> the DB Manager GUI, I cannot create a UNICODE MAPCHAR definition 
> (unfortunately).
> 
> I looked into this because we have an application that is displaying 
> German cities and Öttingen is being displayed at the very end of the 
> list, because the column is (and needs to be) UNICODE.
> 
> It would certainly be nice to be able to address this 
> efficiently at the 
> DB level, w/o needing to add sort logic to the application.

Unfortunately, MaxDB does not support a feature known as 'collations',
and will not support it in the foreseeable future.
(Continue reading)

Permalink | Reply | 501 comments |
headers
John Franklin | 1 Sep 2004 19:04
Favicon

RE: sapdb start up error after a core...

How do I recover from this state???

$DATABASEHOME/indep_prog/bin/dbmcli -d $DATABASE -u dba,dba db_warm
ERR
-24988,ERR_SQL: sql error
-902,Message not available

regards,
johnf

-----Original Message-----
From: John Franklin 
Sent: Monday, August 30, 2004 3:43 PM
To: maxdb <at> lists.mysql.com
Subject: RE: sapdb start up error after a core...

I tried sending the zipped core file but got a failure notice. So
stripped the attachment...

-----Original Message-----
From: John Franklin 
Sent: Monday, August 30, 2004 3:34 PM
To: 'maxdb <at> lists.mysql.com'
Subject: sapdb start up error after a core...

The SAPDB cored out while in use. And now it is failing to come up. How
do I recover?

The sapdb version is 7.3.0.21
(Continue reading)

Permalink | Reply | 1 comment |

Gmane