Re: migrate data 6.5.3 -> 8.3.1 - Mailing list pgsql-hackers

From alexander lunyov
Subject Re: migrate data 6.5.3 -> 8.3.1
Date
Msg-id 48A965E3.8000705@zato.ru
Whole thread Raw
In response to Re: migrate data 6.5.3 -> 8.3.1  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Tom Lane wrote:
> A further suggestion is that you use -d or even -D option on the dump.
> I think there have been some corner-case changes in COPY data format
> since 6.5 days; which might or might not bite you, but why take the
> chance ...


%/usr/local/pgsql/bin/pg_dump -D itt_user > itt_user.dump
Backend sent D message without prior T
Backend sent D message without prior T
...
... (about 2 screens of same messages)
...
Backend sent D message without prior T
Backend sent D message without prior T

and then it hangs.

Then i've tried it with -d option:

%/usr/local/pgsql/bin/pg_dump -d itt_user > itt_user.dump
Killed

I didn't killed pg_dump, so i think it was killed by system after 
pg_dump grows out of some system limit. Size of itt_user.dump is 
something about 2Mb (1974272), last strings in that file are:

INSERT INTO "ip_log" VALUES ('1.1.1.1','user1',300000);
INSERT INTO "ip_log" VALUES ('1.1.1.2','user2',500000);
INSERT INTO "ip_log" VALUES

I crossed my fingers for those dumps i did previously to work.

-- 
alexander lunyov


pgsql-hackers by date:

Previous
From: "Michael Nacos"
Date:
Subject: Re: Overhauling GUCS
Next
From: Peter Eisentraut
Date:
Subject: Compatibility types, type aliases, and distinct types