Since July, has there been any further insight into the postgres backend terminating abnormally? We experienced postgres crashing and then aolserver crashed (with the acs error log reporting the same Message from Postgres...backend died abnormally and possibly corrupted shared memory...), on a site that was pgdump'd and restored on another machine, right on login. -Both machines are running postgres 7.0.2, and same setup of openACS, on FreeBSD.
We did:
on the to-be-backed-up machine:
1. pg_dump -o HRS > hrs.pgdump
on the restore machine:
2. dropdb HRS
3. createdb HRS
4. psql -e < hrs.pgdump
The echoed commands did not show errors as when using cat hrs.pdump |psql HRS (which had errors)
5. Then we restarted nsd
Q: What is the proper way to backup and restore an entire openACS site
thanks.
