On 06/23/2011 12:11 PM, Anoop K wrote: > I am using sqlite3(sqlite-3.7.0.1) with WAL enabled as storage for a > multiprocessing daemon(python). On doing a BULK insert of .5 million rows > each of size 230 bytes in batches of 500 where each batch is a transaction, > following errors happen in other processes which perform (<10) SELECTS and > INSERTS. > > These errors does not happen always and database do function even after > these errors. SELECTS, INSERTS queries after the BULK insert operation do > succeed after the batch operation is completed. > > - file is encrypted or is not a database > - database table is locked > > Sqlite3 configuration > > - 'PRAGMA synchronous=OFF' > - 'PRAGMA journal_mode=wal' > - 'PRAGMA wal_checkpoint=FULL' Rest of the configuration parameters have > have default values > > If I continue BULK inserts with a sleep of 30 sec, after 10+ runs I > see *'database > disk image is malformed' . *After that the table to which BULK inserts were > made does not work. They fail with same error '*database disk image is > malformed'.*. Other tables do respond to queries. > > Any clues on the root cause ? Is this a Sqlite3 WAL bug ?
Do you have a program you can share that exhibits this bug? Dan. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users