Hi Scott
I looked into the release notes of 8.4.2 and found the following fix in
the fix list for 8.4.2 :
------------------------------------------------------------------------
----
Ensure that a cursor's snapshot is not modified after it is created
(Alvaro)
This could lead to a cursor delivering wrong results if later operations
in the same transaction modify the data the cursor is supposed to
return.
------------------------------------------------------------------------
----
The same fix is not included in fix list for postgres-8.1.19 which came
at the same time when postgres-8.4.2 was released i.e 14th Dec.,2009.
Its not there in any of the 8.1 releases after that i.e 8.1.20 and 21.
So , it seems that this fix is not included in 8.1. Do you have any
idea what could be the reason.
Thanks..
Tamanna
-----Original Message-----
From: Scott Marlowe [mailto:scott.marlowe@gmail.com]
Sent: Tuesday, July 13, 2010 1:14 PM
To: tamanna madaan
Cc: pgsql-general@postgresql.org; Tapin Agarwal
Subject: Re: [GENERAL] "attempted to lock invisible tuple" error while
update
On Mon, Jul 12, 2010 at 11:54 PM, tamanna madaan
<tamanna.madan@globallogic.com> wrote:
> Hi Scott
>
> Thanks for your reply . I haven't yet tried updating to latest 8.1.x
version. Was juss googling about this error and came across
> a link discussing the same issue :
>
>
http://groups.google.com/group/pgsql.general/browse_thread/thread/75df15
648bcb502b/10232d1f183a640a?lnk=raot
>
> In this , the problem had occurred on 8.4.1 and a patch
(snapmgr-bugfix-rehash-2.patch ) was provided on 8.4.1 to circumvent
this problem.
>
> This patch worked fine . I was just wondering if this patch
> is backward compatible and can it be merged with 8.1.2
> to get it working. Any suggestions ??
Check the release notes for the 8.1 version that came our around the
same time as 8.4.2? I'd expect if it's a known issue that affected
all versions the fix got put into the older versions as well, unless
it was considered a more dangerous thing to fix there.