Re: load data infile somewhat flaky
"J.O. Aho" <[email protected]> Thu, 19 Aug 2021 15:23:21 +0200
| Newsgroups | comp.databases.mysql |
|---|---|
| Message-ID | <[email protected]> |
On 19/08/2021 14.54, DFS wrote: > On 8/19/2021 6:33 AM, Dr Eberhard Lisse wrote: > >> I find that using tools like CSVQ helps with checking CSV data >> integrity. > > The data came out of a SQLite db and was fairly clean, but encoding > issues meant the SQLite data had 'dupes': > > this field has a question mark ? in it > this field has a question mark � in it As long as it's the same charset used in the sqlite, it shouldn't be too much work to fix it, but a completely different thing is it's a mix. >> Turning off indexes and such during the load and generating them >> afterwards speeds loading of data up significantly. >> >> However, >> >> https://www.google.com/search?q=sqlite+to+mariadb >> >> has 1.5 Million results :-)-O > > > www.google.com/search?q=mariadb+to+sqlite > > has 1.4M results > > SQLite has been very robust for this db of 30M rows. Easy to backup and > administer, never crashed, fast querying... but it's very slow to alter > big tables, and I wanted to see what all the fuss about MySQL/MariaDB > was about. > > It (MariaDB) is a nice dbms, I must say. Especially for no cost/open > source. PostgreSQL is worth to mention (I know shouldn't say it here), also open source and do have some nice features missing in MariaDB/MySQL. -- //Aho