On Jan 19, 2011, at 11:01, Simon Slavin wrote:
> If you do something special to keep your journal file in a different place,
> these other sqlite3 applications won't find it. So they'll just find a
> corrupt database file, and are less likely to be able to figure out how to
> restore to a COMMIT point or a SAVEPOINT.
I understand this concern and think it's a valid point. I can assume
for the purpose of this usage that only tools I provide will be used to access
the DB (I ship a sqlite3 binary since I'm using WAL and I've got users on
CentOS which ships sqlite 1.2 for all I know).
I'm a bit of a proxy of this question. I wrote software that uses
SQLite under some pretty high volumes and I have a user wanting to split stuff
up across multiple filesystems. I already have the ability to do data
partitioning in the application, but the user is wanting to separate the WAL
out as well.
This isn't a question so much about value judgment (I've already argued
that some, though mentioning maintenance tools is helpful there, too). It
comes down to whether reliability of SQLite itself would be reduced if a WAL
existed on a different partition -- whether there are any assumptions WAL makes
that would be invalid across a filesystem boundary.
--
dustin sallings
_______________________________________________
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
