Postgres 6.3 still has problem with the index

Started by Nonameover 27 years ago4 messages
#1Noname
AliE@atlas.com

We were under the impression that the 6.3 resolved the index corruption
issue with regard to large tables(BTP_CHAIN). The problem still exist
and our table indices are still corrupting every night. We are very
desperate and need your help to resolve this issue. Our database size is
about 80M and growing. We have one particular table (User_Account) about
12M, that is heavily accessed and updated. Almost every time the number
of simultaneous access increases, the index on this table corrupts. What
type of information can I present to you to help resolve this issue.

Sincerely

Ali Ebrahimi
Atlas Field Support Manager

#2The Hermit Hacker
scrappy@hub.org
In reply to: Noname (#1)
Re: [HACKERS] Postgres 6.3 still has problem with the index

On Mon, 4 May 1998 AliE@atlas.com wrote:

We were under the impression that the 6.3 resolved the index corruption
issue with regard to large tables(BTP_CHAIN). The problem still exist
and our table indices are still corrupting every night. We are very
desperate and need your help to resolve this issue. Our database size is
about 80M and growing. We have one particular table (User_Account) about
12M, that is heavily accessed and updated. Almost every time the number
of simultaneous access increases, the index on this table corrupts. What
type of information can I present to you to help resolve this issue.

v6.3 or v6.3.2?

I had a problem with this on our server for the longest time, to
the extent that I added code that tells you which index is corrupted (if
it doesn't tell you, you are running an older version)...

v6.3.2, I believe, cleared it up...

Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org

#3Noname
AliE@atlas.com
In reply to: The Hermit Hacker (#2)
RE: [HACKERS] Postgres 6.3 still has problem with the index

Thanks for your response.
We currently are using 6.3. I just looked at the www.postgresql.org and
noticed that under News Flash they have announced the 6.3.2. When I
checked the changes in 6.3.2, I did not notice any fixes with respect to
the Index problem. Is this not Documented? Has it fixed your problem?
Ali Ebrahimi alie@atlas.com

Show quoted text

-----Original Message-----
From: The Hermit Hacker [SMTP:scrappy@hub.org]
Sent: Monday, May 04, 1998 4:48 PM
To: AliE@atlas.com
Cc: pgsql-hackers@postgreSQL.org
Subject: Re: [HACKERS] Postgres 6.3 still has problem with the
index

On Mon, 4 May 1998 AliE@atlas.com wrote:

We were under the impression that the 6.3 resolved the index

corruption

issue with regard to large tables(BTP_CHAIN). The problem still

exist

and our table indices are still corrupting every night. We are very
desperate and need your help to resolve this issue. Our database

size is

about 80M and growing. We have one particular table (User_Account)

about

12M, that is heavily accessed and updated. Almost every time the

number

of simultaneous access increases, the index on this table corrupts.

What

type of information can I present to you to help resolve this issue.

v6.3 or v6.3.2?

I had a problem with this on our server for the longest time, to
the extent that I added code that tells you which index is corrupted
(if
it doesn't tell you, you are running an older version)...

v6.3.2, I believe, cleared it up...

Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary:
scrappy@{freebsd|postgresql}.org

#4The Hermit Hacker
scrappy@hub.org
In reply to: Noname (#3)
RE: [HACKERS] Postgres 6.3 still has problem with the index

On Tue, 5 May 1998 AliE@atlas.com wrote:

Thanks for your response.
We currently are using 6.3. I just looked at the www.postgresql.org and
noticed that under News Flash they have announced the 6.3.2. When I
checked the changes in 6.3.2, I did not notice any fixes with respect to
the Index problem. Is this not Documented? Has it fixed your problem?

I'm not sure if it has just gone away or not, but I haven't seen in since
upgrading to v6.3.2 ...

As for not being documented, possibly not, as it may have been a
'secondary' bug of another bug that was fixed (ie. memory corruption)

Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org