Hello. Long time reader, first time poster here.
I have read the pertinent documentation on this issue but may have
missed some important point. I am puzzled by some results I am seeing.
Apologies in advance for not providing a small code sample that
reproduces the issue, but I think the question should be answerable
without such.
Using sqlite version 3.7.17 (which I know is old but I in my little cube
have no power to change that).
Not using WAL mode.
I have a database with one process (in one thread) writing to it, and
another process (also in a single thread) reading from it only. All
writes are done under BEGIN TRANSACTION IMMEDIATE. Sometimes, an END
TRANSACTION fails with error 5, SQLITE_BUSY. The documentation says
this should not happen:
"The BEGIN IMMEDIATE command might itself return SQLITE_BUSY, but if it
succeeds, then SQLite guarantees that no subsequent operations on the
same database through the next COMMIT will return SQLITE_BUSY."
The read-only process has busy_timeout set to 10000, if that matters.
Am I right that this should not be happening, if indeed my assertions
are correct?
thanks,
Andy
_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
