Linux: 2.0.35
Compaq 450MHz PentiumII with 256MB RAM and 1GB swap
I think I may have dug myself a deep hole. But hopefully someone can help me climb out.
I was vacuuming my database, with analyze on; the vacuum was killed, and I found I could no longer access my DB. Instead, psql would issue an error stating that the DB was not found in pg_database.
Since I had to get the DB back up and running to receive new
data, I saved off the files that represent my primary table (currnt),
and its index, and I ran a script that does a destroyDB/createDB and rebuilds
the schema. I was then operational.
The questions:
1) Is there anyway (in the future) to recover from an aborted
vacuum?
2) Is there anyway to recover the data that is in the table
file I saved?
3) Should all user processes, that are accessing the DB, be
discontinued prior to vacuuming?
4) When I ran my script to recreate the DB and build the tables, it did not create new table files, or update any files in the data/base/<DBname> directory. Since rebuilding, I have added over a million rows to the DB and still there are no files in the directory to represent the tables. Am I missing something here?
--
Steven Wheeler
UNIX Engineering
(918) 292-4119
