Return-Path: postarch
Received: by postgres.Berkeley.EDU (5.61/1.29)
	id AA18015; Wed, 10 Jun 92 15:00:45 -0700
Message-Id: <9206102200.AA18015@postgres.Berkeley.EDU>
From: postarch (Postgres Mailing Archive)
Subject: Re: Unexpected identifier in process_portal
To: postgres@postgres.berkeley.edu
Sender: pg_adm@postgres.berkeley.edu
Reply-To: mer@postgres.berkeley.edu
In-Reply-To: Your message of "Tue, 09 Jun 92 12:49:51 PDT."
             <9206091949.AA03752@postgres.Berkeley.EDU> 
Date: Wed, 10 Jun 92 15:00:31 PDT

you write:

> I had email'd a while back about "Unexpected identifier in process_portal" me
 +ssage.
> Although this happens in a weird variety of circumstance, I have isolated a s
 +imple
> test case. Try the following:
> 
> create Parent(name=text,long=int4)\g
> create Child (short=int2) inherits (Parent) \g
> append Child (name="generic",long=666,short=1)\g
> replace Child (name="generic",long=667,short=1) where Child.long=666\g
> 
> The "replace" command is greeted by: 
>    	Error: No response from the backend, exiting...
> Upon restarting the monitor, the same "replace" command now gives:
> 	Error: Unexpected identifier in process_portal:
> 
> 
> Oddly, a small change to Parent can solve this problem:
> create Parent(long=int4,name=text)\g

I thought we had fixed all of these bad alignment bugs, what version of
postgres are you currently running?  I tried both on my 5000 (Ultrix 4.2a)
and my sparc (SunOs 4.1.2) and couldn't reproduce the bug.

Is there a core dump in the $POSTGRESHOME/data/base/dbname directory? If
so can you send a stack trace?


Jeff Meredith
mer@postgres.berkeley.edu