Critical Upgrade Instruction Bug?
От | phil@Stimpy.netroedge.com |
---|---|
Тема | Critical Upgrade Instruction Bug? |
Дата | |
Msg-id | 20000307190550.F16066@Stimpy.netroedge.com обсуждение исходный текст |
Список | pgsql-general |
I followed the upgrade instructions carefully, and got to this step: psql -e template1 < db.out After a while of waiting (and a lot of stuff scrolling past), it started producing these (after getting into a table with a lot of entries)): [...] query buffer max length of 16384 exceeded query line ignored \smod query buffer max length of 16384 exceeded query line ignored \elevant output from the syslog when the modules are inserted. [...] It seems that pgsql doesn't like all the info shoved down it's throat at once. :'( I used the current (6.5.3) pgdump_all to produce the dump from the old release, and then installed 6.5.3. I checked the drive, and it isn't full. I get better results with: bash /usr/src/pgsql/src/bin/pg_dump/dg_dumpall -d > /tmp/db-backup Instead of: bash /usr/src/pgsql/src/bin/pg_dump/dg_dumpall > /tmp/db-backup W/ each instance being a separate insert, things seem happier (but slower). But, permissions on DB's are lost and mysteriously many tables are empty.... urrrggg!!! Ideas? I'm going from 6.3.? to 6.5.3. Phil -- Philip Edelbrock -- IS Manager -- Edge Design, Corvallis, OR phil@netroedge.com -- http://www.netroedge.com/~phil PGP F16: 01 D2 FD 01 B5 46 F4 F0 3A 8B 9D 7E 14 7F FB 7A
В списке pgsql-general по дате отправления: