A
Aidan Jeffery
Guest
I don't have answers for everything you are seeing, but here are some thoughts. From your description and the errors you are seeing, I am guessing that a) the tables contain non-ASCII characters (so they are doubling in size in the new database) b) the tables have word indexes - on quite large fields In the new database, in case you haven't done so, I would deactivate all the indexes except the primary indexes in each table before loading the data. Rebuild all the indexes once the data is loaded. Other considerations before you start loading the data: 1. Apply an ICU collation to the new databases, e.g., ICU-UCA. If you are upgrading to 11.6, go for the more recent version (e.g. ICU_48-UCA). 2. Ensure that the UTF-8 word-break table is applied to the new databases. If they are created from DLC/prolang/utf/empty, then that will be the case.
Continue reading...
Continue reading...