Re: [BUGS] BUG #1727: Corrupt Zip file?

2005-06-26 Thread Magnus Hagander
> The following bug has been logged online:
> 
> Bug reference:  1727
> Logged by:  Robert Zahm
> Email address:  [EMAIL PROTECTED]
> PostgreSQL version: 8.0.3
> Operating system:   Windows 2000
> Description:Corrupt Zip file?
> Details: 
> 
> I an unable to unzip the .zip file for the 8.0.3 download.  I 
> receive the following error from winzip: The CRC is 9822df6f, 
> it should be c1ddd3d1.
> 
> I was able to download and unzip 8.0 and 8.0.2 without any 
> issues.  I downloaded the 8.0.3 zip file 3 times.

Which mirror did you download this file from? If you've always tried the
same, please try another one.

A bad version of this file was initially uploaded at the 8.0.3 release,
but it was replaced with a correct one a long time ago. Perhaps there is
a mirror out there that hasn't picked up the change for some reason.

//Magnus

---(end of broadcast)---
TIP 9: In versions below 8.0, the planner will ignore your desire to
   choose an index scan if your joining column's datatypes do not
   match


[BUGS] Works as SELECT but not as a VIEW

2005-06-26 Thread Robert Creager

This doesn't seem right:

CREATE VIEW inci AS
SELECT date, inci_id, protocol, a.ip_addr, b.ip_addr, spt, dpt
FROM  (SELECT addr_id, ip_addr FROM addrs) AS a
 JOIN incident ON incident.source_addr_id = a.addr_id 
 JOIN (SELECT addr_id, ip_addr FROM addrs) 
   AS b ON incident.destination_addr_id = b.addr_id
 JOIN protocols USING( protocol_id );

Produces: ERROR:  column "ip_addr" duplicated

Yet the select on it's own works just fine...  I've attached the database
definition.

 version

---
--
 PostgreSQL 8.0.2 on i686-pc-linux-gnu, compiled by GCC gcc (GCC) 3.3.2
(Mandrake Linux 10.0 3.3.2-6mdk)


Thanks,
Rob

-- 
 12:31:42 up 24 days, 14:10,  6 users,  load average: 3.08, 2.63, 2.63
Linux 2.6.5-02 #8 SMP Mon Jul 12 21:34:44 MDT 2004


test.sql
Description: Binary data


pgplqDYerWTQO.pgp
Description: PGP signature


Re: [BUGS] Works as SELECT but not as a VIEW

2005-06-26 Thread Tom Lane
Robert Creager <[EMAIL PROTECTED]> writes:
> This doesn't seem right:

> CREATE VIEW inci AS
> SELECT date, inci_id, protocol, a.ip_addr, b.ip_addr, spt, dpt
> FROM  (SELECT addr_id, ip_addr FROM addrs) AS a
>  JOIN incident ON incident.source_addr_id =3D a.addr_id=20
>  JOIN (SELECT addr_id, ip_addr FROM addrs)=20
>AS b ON incident.destination_addr_id =3D b.addr_id
>  JOIN protocols USING( protocol_id );

> Produces: ERROR:  column "ip_addr" duplicated

Why doesn't it seem right?  You're asking it to generate a view with
two columns both named ip_addr.  Use "AS" to rename one or both.

regards, tom lane

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]


Re: [BUGS] Works as SELECT but not as a VIEW

2005-06-26 Thread Robert Creager
When grilled further on (Sun, 26 Jun 2005 15:18:47 -0400),
Tom Lane <[EMAIL PROTECTED]> confessed:

> Robert Creager <[EMAIL PROTECTED]> writes:
> > This doesn't seem right:
> 
[ snip embarrassing mistake ]
> 
> Why doesn't it seem right?  You're asking it to generate a view with
> two columns both named ip_addr.  Use "AS" to rename one or both.
> 

So the SELECT can manage this bot not the VIEW?  Thanks for catching my mistake.

Cheers,
Rob

-- 
 13:29:26 up 24 days, 15:07,  6 users,  load average: 1.90, 2.31, 2.78
Linux 2.6.5-02 #8 SMP Mon Jul 12 21:34:44 MDT 2004


pgpkIQjVQ9rqr.pgp
Description: PGP signature


Re: [BUGS] Works as SELECT but not as a VIEW

2005-06-26 Thread Tom Lane
Robert Creager <[EMAIL PROTECTED]> writes:
> So the SELECT can manage this bot not the VIEW?

The spec constrains tables (and views) not to have duplicate column
names, but it does not say that about SELECT output lists.

regards, tom lane

---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faq


Re: [BUGS] can't startup PG-8.0.3

2005-06-26 Thread Ruslan A Dautkhanov




Hello !

Tom Lane wrote:

  
I have a command
su - pgsql -c '/usr/local/pgsql/bin/postmaster -d 5 -D 
/usr/local/pgsql/data'

  
  
Shouldn't there be an & or something in there?  Seems like that will
leave the su command blocked waiting for the postmaster to exit,
which of course it should not.

  

I try to use &, this did not help.

  
  
I even tries to use ktrace(1) :
ktrace su - pgsql -c '/usr/local/pgsql/bin/postmaster -d 5 -D 
/usr/local/pgsql/data'

  
  
That won't tell you anything --- the trace will only trace what "su" does.
Better to su to pgsql and ktrace the postmaster directly.
  

Yes, you're right. The ktrace dump for postmaster process:
su - pgsql -c 'ktrace /usr/local/pgsql/bin/postmaster -i -d 5 -D
/usr/local/pgsql/data' :

...
 86611 postgres RET   write 44/0x2c
 86611 postgres CALL  gettimeofday(0xbfbfd5a8,0)
 86611 postgres RET   gettimeofday 0
 86611 postgres CALL  getpid
 86611 postgres RET   getpid 86611/0x15253
 86611 postgres CALL  sendto(0x5,0xbfbfda80,0x60,0,0,0)
 86611 postgres GIO   fd 5 wrote 96 bytes
   "<135>Jun 27 11:02:02 postgres[86611]: [569-1] %DEBUG: 
-"
 86611 postgres RET   sendto 96/0x60
 86611 postgres CALL  write(0x2,0x82eb128,0x33)
 86611 postgres GIO   fd 2 wrote 51 bytes
   "%DEBUG:  -
   "
 86611 postgres RET   write 51/0x33
 86611 postgres CALL  fork
 86611 postgres RET   fork 86612/0x15254
 86611 postgres CALL  exit(0)


I'll keep try to startup PostgreSQL 8.0.3...  Thanks


--
Ruslan A Dautkhanov




smime.p7s
Description: S/MIME Cryptographic Signature


Re: [BUGS] can't startup PG-8.0.3

2005-06-26 Thread Ruslan A Dautkhanov




Tom Lane wrote:

  Ruslan A Dautkhanov <[EMAIL PROTECTED]> writes:
  
  
I have experience with PostgreSQL 7.x, and I can't figure out why this 
happen. PG didn't
say anything meaningful, that can point me why he did not start.

  
  
That's fairly unusual --- I can't think of any places where the
postmaster will exit without printing *some* complaint.
  


# ./pg_controldata /usr/local/pgsql/data
    pg_control version number:    74
    Catalog version number:   200411041
    Database system identifier:   4786296063829923563
    Database cluster state:   in production
    pg_control last modified: среда, 22 июня 2005 г.
03:12:28
    Current log file ID:  0
    Next log file segment:    13
    Latest checkpoint location:   0/C1DB058
    Prior checkpoint location:    0/C0E47BC
    Latest checkpoint's REDO location:    0/C1DB058
    Latest checkpoint's UNDO location:    0/0
    Latest checkpoint's TimeLineID:   1
    Latest checkpoint's NextXID:  133213
    Latest checkpoint's NextOID:  492366
    Time of latest checkpoint:    среда, 22 июня 2005 г.
03:12:28
    Database block size:  8192
    Blocks per segment of large relation: 131072
    Bytes per WAL segment:    16777216
    Maximum length of identifiers:    64
    Maximum number of function arguments: 32
    Date/time type storage:   floating-point numbers
    Maximum length of locale name:    128
    LC_COLLATE:   ru_RU.KOI8-R
    LC_CTYPE: ru_RU.KOI8-R

Please note different values in "Latest checkpoint location" and "Prior
checkpoint location" - is this mean
that DB shutted down abnormally and some uncommited data exists? Maybe
Postgres can't commit
this changes?   

Btw, before this version I use PG 8.0.2 and after this I install 8.0.3,
that can't startup... :/ 
Maybe 8.0.3 conflict with DB created in 8.0.2 ?

Thanks a lot for your comments.

--
Ruslan A Dautkhanov





smime.p7s
Description: S/MIME Cryptographic Signature


[BUGS] problem with unexpected error

2005-06-26 Thread satish reddy

Hi,
I'm using postgreSQL with vb.net is working fine. But some times i got the error is "ERROR: no error information is available". I'm using 8.1 version and odbc 9.x version. Please give me some solution to avoid this error
 
Thanks in advance
 
Satish
		Yahoo! Sports 
Rekindle the Rivalries. Sign up for Fantasy Football




Re: [BUGS] can't startup PG-8.0.3

2005-06-26 Thread Tom Lane
Ruslan A Dautkhanov <[EMAIL PROTECTED]> writes:
> Please note different values in "Latest checkpoint location" and "Prior 
> checkpoint location" - is this mean
> that DB shutted down abnormally and some uncommited data exists?

No; indeed, if they were the same *then* I would worry.

regards, tom lane

---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
   subscribe-nomail command to [EMAIL PROTECTED] so that your
   message can get through to the mailing list cleanly