After call commit sql , data can not flush to disk

Question

I use berkey db which support sql . It's version is db-5.1.19.
1, Open a database.
2. Create a table.
3. exec "begin;" sql
4. exec sql which is insert record into table
5. exec "commit;" sql
6. copy database file (SourceDB_912_1.db and SourceDB_912_1.db-journal) to Local Disk of D, then use a tool of dbsql to open the database.
7. use select sql to check data, there is no record in table.
1
sqlite3 * m_pDB;
int nRet = sqlite3_open_v2(strDBName.c_str(), & m_pDB,SQLITE_OPEN_READWRITE | SQLITE_OPEN_CREATE,NULL);
2
string strSQL="CREATE TABLE [TBLClientAccount] ( [ClientId] CHAR (36), [AccountId] CHAR (36) );";
char * errors;
nRet = sqlite3_exec(m_pDB, strSQL.c_str(), NULL, NULL, &errors);
3
nRet = sqlite3_exec(m_pDB, "begin;", NULL, NULL, &errors);
4
nRet = sqlite3_exec(m_pDB, "INSERT INTO TBLClientAccount (ClientId,AccountId) VALUES('dd','ddd'); ", NULL, NULL, &errors);
5
nRet = sqlite3_exec(m_pDB, "commit;", NULL, NULL, &errors);
Edited by: 887973 on Sep 27, 2011 11:15 PM

Answer 1

Hi,
Here is a simple test case program I used based on your description:
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "sqlite3.h"
int error_handler(sqlite3*);
int main()
     sqlite3 *m_pDB;
     const char *strDBName = "C:/SRs/OTN Core 2290838 - after call commit sql , data can not flush to disk/SourceDB_912_1.db";
     char * errors;
     sqlite3_open_v2(strDBName, &m_pDB, SQLITE_OPEN_READWRITE | SQLITE_OPEN_CREATE, NULL);
     error_handler(m_pDB);
     sqlite3_exec(m_pDB, "CREATE TABLE [TBLClientAccount] ( [ClientId] CHAR (36), [AccountId] CHAR (36) );", NULL, NULL, &errors);
     error_handler(m_pDB);
     sqlite3_exec(m_pDB, "begin;", NULL, NULL, &errors);
     error_handler(m_pDB);
     sqlite3_exec(m_pDB, "INSERT INTO TBLClientAccount (ClientId,AccountId) VALUES('dd','ddd'); ", NULL, NULL, &errors);
     error_handler(m_pDB);
     sqlite3_exec(m_pDB, "commit;", NULL, NULL, &errors);
     error_handler(m_pDB);
     //sqlite3_close(m_pDB);
     //error_handler(m_pDB);
int error_handler(sqlite3 *db)
     int err_code = sqlite3_errcode(db);
     switch(err_code) {
     case SQLITE_OK:
     case SQLITE_DONE:
     case SQLITE_ROW:
          break;
     default:
          fprintf(stderr, "ERROR: %s. ERRCODE: %d.\n", sqlite3_errmsg(db), err_code);
          exit(err_code);
     return err_code;
}Than I copied the SourceDB_912_1.db database and the SourceDB_912_1.db-journal directory containing the environment files (region files, log files) to D:\, opened the database using the "dbsql" command line tool, and queried the table; the data is there:
D:\bdbsql-dir>ls -al
-rw-rw-rw-   1 acostach 0 32768 2011-10-12 12:51 SourceDB_912_1.db
drw-rw-rw-   2 acostach 0     0 2011-10-12 12:51 SourceDB_912_1.db-journal
D:\bdbsql-dir>C:\BerkeleyDB\db-5.1.19\build_windows\Win32\Debug\dbsql SourceDB_912_1.db
Berkeley DB 11g Release 2, library version 11.2.5.1.19: (August 27, 2010)
Enter ".help" for instructions
Enter SQL statements terminated with a ";"
dbsql> .tables
TBLClientAccount
dbsql> .schema TBLClientAccount
CREATE TABLE [TBLClientAccount] ( [ClientId] CHAR (36), [AccountId] CHAR (36) );
dbsql> select * from TBLClientAccount;
dd|dddI do not see where the issue is. The data can be successfully retrieved, it is present in the database.
Could you try putting in the sqlite3_close() call and see if you still get the error?
Did you remove the __db.* files from the SourceDB_912_1.db-journal directory?
Did you use PRAGMA synchronous, and if so, what is the value you set?
If this is still an issue for you, please describe in more detail the exact steps needed to get this reproduced and provide a simple stand-alone test case program that reproduces it.
Regards,
Andrei