Thread: Errors when running vacuumdb...
when I run vacuumdb -a -v, I get: NOTICE: Rel pg_attribute: Uninitialized page 1 - fixing ... more lines for page 2 through 50 ... NOTICE: Rel pg_attribute: Uninitialized page 51 - fixing Any idea what causes this? -- Dominic J. Eidson "Baruk Khazad! Khazad ai-menu!" - Gimli ------------------------------------------------------------------------------- http://www.the-infinite.org/ http://www.the-infinite.org/~dominic/
"Dominic J. Eidson" <sauron@the-infinite.org> writes: > when I run vacuumdb -a -v, I get: > NOTICE: Rel pg_attribute: Uninitialized page 1 - fixing > ... more lines for page 2 through 50 ... > NOTICE: Rel pg_attribute: Uninitialized page 51 - fixing > Any idea what causes this? This is not good --- VACUUM is trying to recover from finding all-zero pages in your table. I would expect that the messages wouldn't recur after you vacuum the table once; if they are repeatable then something really strange is going on. But the real question is why such pages are in the file in the first place. What PG version are you running? Have you had any crashes or other strange events lately? regards, tom lane
On Tue, 2 Apr 2002, Tom Lane wrote: > "Dominic J. Eidson" <sauron@the-infinite.org> writes: > > when I run vacuumdb -a -v, I get: > > NOTICE: Rel pg_attribute: Uninitialized page 1 - fixing > > ... more lines for page 2 through 50 ... > > NOTICE: Rel pg_attribute: Uninitialized page 51 - fixing > > > Any idea what causes this? > > This is not good --- VACUUM is trying to recover from finding all-zero > pages in your table. I would expect that the messages wouldn't recur > after you vacuum the table once; if they are repeatable then something I just re-ran vacuumdb -a -z, and the same errors ocurred. > really strange is going on. But the real question is why such pages > are in the file in the first place. No idea. :) I dropped the DB, and recreated - ran vacuumdb against it, and get no errors. I'm currently in the process of re-populating the DB (should take a few hours) - but will re-run vacuumdb -a -z, once it's done. > What PG version are you running? Have you had any crashes or other > strange events lately? sauron=# select version(); version ------------------------------------------------------------- PostgreSQL 7.2 on i686-pc-linux-gnu, compiled by GCC 2.95.3 A few not-so-clean shutdowns, maybe - but no crashes to my knowledge. Other than a hard drive crash last wednesday (subsequently reinstalled system, dumped all data, upgraded to 7.2 (was 7.1), and reloaded the data). IOW, the drive crash should have preciously little to do with the above errors.. We'll see what happens once it's done loading data. -- Dominic J. Eidson "Baruk Khazad! Khazad ai-menu!" - Gimli ------------------------------------------------------------------------------- http://www.the-infinite.org/ http://www.the-infinite.org/~dominic/
On Tue, 2 Apr 2002, Dominic J. Eidson wrote: > On Tue, 2 Apr 2002, Tom Lane wrote: > > "Dominic J. Eidson" <sauron@the-infinite.org> writes: > > > when I run vacuumdb -a -v, I get: > > > NOTICE: Rel pg_attribute: Uninitialized page 1 - fixing > > > ... more lines for page 2 through 50 ... > > > NOTICE: Rel pg_attribute: Uninitialized page 51 - fixing > > > Any idea what causes this? > > This is not good --- VACUUM is trying to recover from finding all-zero > > pages in your table. I would expect that the messages wouldn't recur > > after you vacuum the table once; if they are repeatable then something > > I just re-ran vacuumdb -a -z, and the same errors ocurred. [snip] > No idea. :) I dropped the DB, and recreated - ran vacuumdb against it, and > get no errors. I'm currently in the process of re-populating the DB > (should take a few hours) - but will re-run vacuumdb -a -z, once it's > done. Database fully populated. Re-ran vacuumdb -a -z , and it runs clean. -- Dominic J. Eidson "Baruk Khazad! Khazad ai-menu!" - Gimli ------------------------------------------------------------------------------- http://www.the-infinite.org/ http://www.the-infinite.org/~dominic/