From: Achilleas Mantzios <achill@matrix.gatewaynet.com>
Organization: Dynacom Tankers Mgmt
To: pgsql-admin@postgresql.org
Subject: pg_dump: Error message from server: ERROR: compressed data is corrupt
Date: Fri, 23 Apr 2010 13:54:42 +0200
User-Agent: KMail/1.9.7
MIME-Version: 1.0
Content-Type: text/plain;
  charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Message-Id: <201004231454.43104.achill@matrix.gatewaynet.com>

Hello,
just coming back from a rescue marathon on this remote server i was telling=
 you.
As i said, the last problem was while doing a=20
pg_dump dynacom=20
(dynacom is my db'd name)
i kept getting=20

pg_dump: SQL command failed
pg_dump: Error message from server: ERROR:  compressed data is corrupt
pg_dump: The command was: COPY public.mail_entity (msgno, entno, entparentn=
o....

so i tried dumping the individual mail_entity table, which resulted in the =
following errors:
pg_dump -t mail_entity > /dev/null
pg_dump: SQL command failed
pg_dump: Error message from server: ERROR:  missing chunk number 2 for toas=
t value 133502683
pg_dump: The command was: COPY public.mail_entity (msgno, entno, entparentn=
o, entdate, entrecvdate, entsendername, entsenderaddr, entmess
ageid, en;
srv:~>

=46rom the 76000 rows in that table i found the damaged columns by using th=
e following technique=20
(which requires O(log(n)) selects btw lol!!!)
SELECT msgno,entno,entdate from mail_entity where coalesce(entdata::text,''=
)||coalesce(enttext,'') like '%foo%' and msgno/10000 =3D 0;
SELECT msgno,entno,entdate from mail_entity where coalesce(entdata::text,''=
)||coalesce(enttext,'') like '%foo%' and msgno/10000 =3D 1;
=2E... until i found the correct subset of 10000 rows
SELECT msgno,entno,entdate from mail_entity where coalesce(entdata::text,''=
)||coalesce(enttext,'') like '%foo%' and msgno/10000 =3D 7;
then
SELECT msgno,entno,entdate from mail_entity where coalesce(entdata::text,''=
)||coalesce(enttext,'') like '%foo%' and msgno/1000 =3D 70;
SELECT msgno,entno,entdate from mail_entity where coalesce(entdata::text,''=
)||coalesce(enttext,'') like '%foo%' and msgno/1000 =3D 71;
=2E..
and so on untill i found the two (it was two of them) bad rows.
i did entdata=3D'' for those two rows, and the selects could be done withou=
t issues.

however, after that, the pg_dump -t reverted back to the aforementioned err=
or.
Now doing this gives:
srv:~> pg_dump -t mail_entity > /dev/null
pg_dump: SQL command failed
pg_dump: Error message from server: ERROR:  compressed data is corrupt
pg_dump: The command was: COPY public.mail_entity (msgno, entno, entparentn=
o, entdate, entrecvdate, entsendername, entsenderaddr, entmessageid, en;
srv:~> =20

Then i did

# CREATE TABLE mail_entity2 AS SELECT * FROM mail_entity;

which went fine

but, for some crazy reason, pg_dump on mail_entity2 also results to an erro=
r:
srv:~> pg_dump -t mail_entity2 > /dev/null
pg_dump: SQL command failed
pg_dump: Error message from server: ERROR:  compressed data is corrupt
pg_dump: The command was: COPY public.mail_entity2 (msgno, entno, entparent=
no, entdate, entrecvdate, entsendername, entsenderaddr, entmessageid, e;
srv:~>  =20

Is there anything i could do to take this very much wanted dump?

=2D-=20
Achilleas Mantzios