Обсуждение: dumping tables from badly damaged db

Recently I had a problem where a system crash scribbed on some directories,
which landed a bunch of files, including a few of the system table files for
one of my databases, in lost+found along with a zillion other files.

I might be able to find the file for this table/index in lost+found, but how
do I know what name to give it in /var/lib/postgres/...? These files are
named after OID's which I can't find because I can't connect to the
database.

If I can't get the file back, is there any way I can dump the data out of
the remaining tables?

Currently, I can't connect to the database with the following error:

psql: FATAL 1:  cannot open pg_class_relname_index: No such file or directory

I suspect that's very bad, considering the centrality of the pg_class
table...

--
Brian Ristuccia
bristucc@sw.starentnetworks.com

Brian Ristuccia <bristucc@sw.starentnetworks.com> writes:
> Recently I had a problem where a system crash scribbed on some directories,
> which landed a bunch of files, including a few of the system table files for
> one of my databases, in lost+found along with a zillion other files.

Ugh.

> I might be able to find the file for this table/index in lost+found, but how
> do I know what name to give it in /var/lib/postgres/...?

I can't think of any reasonably simple way to identify the files by
content (this might be something to try to fix in future, but for now
you're stuck).  Best idea I can think of is to examine "od -c" dumps
and try to intuit which file is which.

> Currently, I can't connect to the database with the following error:
> psql: FATAL 1:  cannot open pg_class_relname_index: No such file or directory

You might be able to get past this by starting a standalone postgres
with the -P command-line option (ignore system indexes).  If so, try
"select relname, relfilenode from pg_class".  With luck that will give
you a list of which file name is needed for each table.  I'd not counsel
trying to do more than that in the standalone backend until you've
gotten at least the tables put back together.

You do not need to try very hard to recreate the indexes --- you can use
REINDEX to rebuild them.

Good luck!

            regards, tom lane

Brian Ristuccia <bristucc@sw.starentnetworks.com> writes:
> The standalone backend errors out with:
> FATAL 1:  _mdfd_getrelnfd: cannot open relation pg_trigger: No such file or
> directory

Well, if you can identify which of the lost+found files is pg_trigger,
you can move it back into place and then try again.  (Look for trigger
names in the od -c dump...)  All the system tables have fixed names
(relfilenode values) which you can determine by consulting another
database of the same PG version.  pg_trigger is 16412 in 7.3, for
instance.

Lather, rinse, repeat until it comes up ...

> My week-old backups are starting to look more and more attractive.

I didn't say this was going to be painless.

            regards, tom lane

On Fri, Oct 31, 2003 at 02:05:57PM -0500, Tom Lane wrote:
>
> You might be able to get past this by starting a standalone postgres
> with the -P command-line option (ignore system indexes).  If so, try
> "select relname, relfilenode from pg_class".  With luck that will give
> you a list of which file name is needed for each table.  I'd not counsel
> trying to do more than that in the standalone backend until you've
> gotten at least the tables put back together.
>

The standalone backend errors out with:

FATAL 1:  _mdfd_getrelnfd: cannot open relation pg_trigger: No such file or
directory

I suspect my troubles may be more severe than just a missing index... My
week-old backups are starting to look more and more attractive.

Any other hints?

Thanks.

-Brian

On Sun, Nov 02, 2003 at 01:45:39PM -0500, Brian Ristuccia wrote:
>
> pg_dump: NOTICE:  RelationBuildDesc: can't open deprecated_data1: No such file or directory
> pg_dump: NOTICE:  RelationBuildDesc: can't open unimportant_stuff: No such file or directory
> pg_dump: handler procedure for procedural language plpgsql not found
>
> I can't drop the tables from psql:
>
> drop TABLE deprecated_data1 ;
> ERROR:  cannot open deprecated_data1: No such file or directory
>

Ok. I managed to drop these in the standalone backend with some hackery:

backend> drop table useless_junk
NOTICE:  RelationBuildDesc: can't open useless_junk: No such file or directory
ERROR:  cannot open useless_junk: No such file or directory

So I'd have to go select on pg_class and find the relfilenode, go touch that
file so it exists. Try the drop again. Drop any indexes which don't exist.
Find the relfilenode for the associated pg_toast table and go touch that
file too. Not a lot of fun, but the destroyed tables are gone now.

> Am I likely to have success by performing surgery on the pg_table relation
> in order to remove these destroyed tables?

A simple DROP LANGUAGE did the trick.

I can now run pg_dump - I feel much better.

--
Brian Ristuccia

On Fri, Oct 31, 2003 at 02:45:57PM -0500, Tom Lane wrote:
> Lather, rinse, repeat until it comes up ...
>

Ok. That excercise was certainly no fun. For the various system tables I was
missing, I managed to either find them in lost+found, or I copied the
version from template0. The files for some relatively unimportant tables
were irrecoverably destroyed.

Of course, I realize that my database is probably now in a percarious state.
I want to run a pg_dump, drop the db, and then rebuild into a fresh
database. Except for selects on the destroyed tables, my database works as
normal, so I'm confident I can get my data out.

But pg_dump refuses to run. It complains about the destroyed tables and also
something about plpgsql:

pg_dump: NOTICE:  RelationBuildDesc: can't open deprecated_data1: No such file or directory
pg_dump: NOTICE:  RelationBuildDesc: can't open unimportant_stuff: No such file or directory
pg_dump: handler procedure for procedural language plpgsql not found

I can't drop the tables from psql:

drop TABLE deprecated_data1 ;
ERROR:  cannot open deprecated_data1: No such file or directory

Am I likely to have success by performing surgery on the pg_table relation
in order to remove these destroyed tables? Any hints on what I should do
about the procedural language issue?

--
Brian Ristuccia