mariadb/mysql-test
unknown 05b709fc77 WL#1339 "Add per account max_user_connections limit (maximum number
of concurrent connections for the same account)"

Added support of account specific max_user_connections limit. Made all
user limits to be counted per account instead of the old behavior,
which was per user/host accounting. Added option which enables the old
behavior. Added testing of these to the test suite.

(After review version).


client/mysqltest.c:
  Extended mysqltest to be able to handle connect() statements for which
  error is expected.
  
  - Added replace_dynstr_append() utility function.
  - added connect_n_handle_errors() function which connects with server
    without retries and handles errors as if "connect" was usual
    statement.
  - do_connect(): added ability to handle connects which are expected
    to return an error.
  - run_query_normal(): Moved all expected-error-handling code to separate
    normal_handle_error()/normal_handle_no_error() functions to be able
    to reuse them in connect_n_handle_errors().
mysql-test/r/grant.result:
  Fixed test results since one more column to mysql.user was added.
mysql-test/r/system_mysql_db.result:
  Fixed test results since one more column to mysql.user was added.
scripts/mysql_create_system_tables.sh:
  Added max_user_connections column to mysql.user table for storing
  new maximal concurrent connections per account limit.
scripts/mysql_fix_privilege_tables.sql:
  Added max_user_connections column to mysql.user table for storing
  new maximal concurrent connections per account limit.
sql/lex.h:
  Added MAX_USER_CONNECTIONS symbol used for specifying
  maximum number of concurrent connections per account.
sql/mysql_priv.h:
  Added declaration of opt_old_style_user_limits variable which is defined
  in sql/mysqld.cc used in sql/sql_parse.cc.
sql/mysqld.cc:
  Added "old-style-user-limits" option which forces user limits to behave
  in old way i.e. to be counted per user/host pair instead of per account.
  Added comment describing mqh_used variable.
sql/set_var.cc:
  Added sys_var_max_user_conn class which implements support for the new 
  behavior of max_user_connections variable. Now the global instance of
  this variable holds default maximum number of concurrent connections per
  account (as it was before) and the session instance gives read-only
  access to account-specific version of this limit.
sql/set_var.h:
  Added sys_var_max_user_conn class which implements support for the new 
  behavior of max_user_connections variable. Now the global instance of
  this variable holds default maximum number of concurrent connections per
  account (as it was before) and the session instance gives read-only
  access to account-specific version of this limit.
sql/sql_acl.cc:
  Added support for account-specific MAX_USER_CONNECTIONS limit.
  Tweaked USER_RESOURCES and their handling for better clarity.
sql/sql_parse.cc:
  Added support for account-specific MAX_USER_CONNECTIONS (maximum number
  of concurrent connections per account) limit. Changed default behavior
  of all user limits to be per account instead of per user+host.
  '--old-style-user-limits' option was added to enable the old behavior.
  Made maximum number of connections per hour to be independant on the
  value of global max_user_connections variable.
sql/sql_yacc.yy:
  Added support of new MAX_USER_CONNECTIONS limit to grammar.
  Renamed USER_RESOURCES::connections member to conn_per_hour and bits
  member to specified_limits. Also enum is used instead of naked numbers
  when we are working with specified_limits.
sql/structs.h:
  USER_RESOURCES struct:
   - Added user_conn member to store the maximum number of concurrent 
     connections for an account. Renamed connections member to 
     conn_per_hour for less ambiguity.
   - Renamed member 'bits' to 'specified_limits' for the sake of clarity.
     The member was used as a flag indicating which limits were mentioned
     in GRANT clause.
   - Added comments.
  USER_CONN struct:
   - Removed unused user_len member.
   - Added comments.
2004-12-29 20:30:37 +03:00
..
include merged 2004-11-20 18:36:41 +01:00
misc
ndb ndbcluster.sh: 2004-11-11 11:08:37 +00:00
r WL#1339 "Add per account max_user_connections limit (maximum number 2004-12-29 20:30:37 +03:00
std_data merge with 4.0 2004-11-12 11:17:53 +02:00
t WL#1339 "Add per account max_user_connections limit (maximum number 2004-12-29 20:30:37 +03:00
create-test-result
fix-result
init_db.sql replaced init_db.sql 2004-11-10 15:03:59 +05:00
install_test_db.sh added 2 masters for ndb tests 2004-09-28 17:04:30 +00:00
Makefile.am merge with 4.0 2004-11-12 11:17:53 +02:00
my_manage.c mysql_test_run_new.c, my_manage.h, my_manage.c: 2004-11-09 20:21:37 +01:00
my_manage.h mysql_test_run_new.c, my_manage.h, my_manage.c: 2004-11-09 20:21:37 +01:00
mysql-test-run.sh Resolved merge from 4.1 -> 5.0 (mainly Guilhem's work with mysqldump) 2004-11-11 18:58:57 -08:00
mysql_test_run_new.c mysql_test_run_new.c, my_manage.h, my_manage.c: 2004-11-09 20:21:37 +01:00
README
README.gcov
resolve-stack

This directory contains a test suite for mysql daemon. To run
the currently existing test cases, simply execute ./mysql-test-run in
this directory. It will fire up the newly built mysqld and test it.

If you want to run the test with a running MySQL server use the --external
option to mysql-test-run.

Note that you do not have to have to do make install, and you could
actually have a co-existing MySQL installation - the tests will not
conflict with it.

All tests must pass. If one or more of them fail on your system, please
read the following manual section of how to report the problem:

http://dev.mysql.com/doc/mysql/en/MySQL_test_suite.html


You can create your own test cases. To create a test case:

 xeamacs t/test_case_name.test

 in the file, put a set of SQL commands that will create some tables,
 load test data, run some queries to manipulate it.

 We would appreciate if the test tables were called t1, t2, t3 ... (to not
 conflict too much with existing tables).

 Your test should begin by dropping the tables you are going to create and
 end by dropping them again.  This will ensure that one can run the test
 over and over again.
 
 If you are using mysqltest commands (like result file names) in your
 test case you should do create the result file as follows:

 mysql-test-run --record test_case_name

 or

 mysqltest --record < t/test_case_name.test

 If you only have a simple test cases consistent of SQL commands and comments
 you can create the test case one of the following ways:

 mysql-test-run --record test_case_name

 mysql test < t/test_case_name.test > r/test_case_name.result

 mysqltest --record --record-file=r/test_case_name.result < t/test_case_name.test

 When this is done, take a look at r/test_case_name.result
 - If the result is wrong, you have found a bug;  In this case you should
   edit the test result to the correct results so that we can verify
   that the bug is corrected in future releases.

To submit your test case, put your .test file and .result file(s) into
a tar.gz archive, add a README that explains the problem, ftp the 
archive to ftp://support.mysql.com/pub/mysql/secret/ and send  a mail
to bugs@lists.mysql.com