Re: 8.2.6 > 8.3 blows up

[Karl Denninger <karl@xxxxxxxxxxxxx>]

A whole host of them, mostly about bad data formats in some of the table data. I suspect the underlying problem is that something got mangled in the table creates.

I'm setting up on a different box as my attempt to create a second instance failed horribly - compiling with a different prefix and port number appears to work but when you do an "initdb" it blows up with a complaint about not being able to create the semaphores and shared segment.  It appears the port number is used for the SEMID and SHMID prefixes, and those for some reason are not getting reset (it may be that the change in configure requires a gmake clean; not sure)

In any event I have another machine and will get something more detailed ASAP - I will also try the "restore" program and see if that works.

Karl Denninger (karl@xxxxxxxxxxxxx)
http://www.denninger.net

Scott Marlowe wrote:

On Sun, Mar 2, 2008 at 1:41 PM, Karl Denninger <karl@xxxxxxxxxxxxx> wrote:
  

Ugh.

 I am attempting to move from 8.2.6 to 8.3, and have run into a major
 problem.

 The build goes fine, the install goes fine, the pg_dumpall goes fine.

 However, the reload does not.  I do the initdb and then during the
 reload I get thousands of errors, apparently from table data which is
 perfectly ok in the tables, but pukes up a hairball when attempted to be
 reloaded.
    

So what error message?

---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
       subscribe-nomail command to majordomo@xxxxxxxxxxxxxx so that your
       message can get through to the mailing list cleanly


%SPAMBLOCK-SYS: Matched [hub.org+], message ok