The test repeatedly blocks and unblocks a master which goes unnoticed by
the monitor due to the 10 second read timeouts in the monitor
configuration in the longblob template. The replication template uses the
default timeouts which makes the test actually exercise the functionality
that it is intended to test.
By always starting the session shutdown process by stopping the client
DCB, the manipulation of the session state can be removed from the backend
protocol modules and replaced with a fake hangup event.
Delivering this event via the core allows the actual dcb_close call on the
client DCB to be done only when the client DCB is being handled by a
worker.
Directly closing the client DCB in the backend protocol modules is not
correct anymore as the state of the session doesn't change when the client
DCB is closed. By propagating the shutdown of the session with a fake
hangup to the client DCB, the closing of the DCB is done only once.
Added debug assertions that make sure all DCBs are closed only
once. Removed redundant code in the backend protocol error handling code.
The token skipping function did not check for a period or an opening
parenthesis when parsing the test. Also fixed a debug assertion when only
NULL values were inserted.
The test now also creates TIME type values and checks that they are
converted correctly. Also added NULL value tests for all values and made
required adjustments to the code.
In debug mode, when scanning the server id from a string, check that resulting
number is 32bit. Also, when querying the server id, query the global version.
Now, if a super user modifies the server id the monitor will notice it.
Server id:s in gtid:s are handled similarly.
When a double close is detected in a debug build, a debug assertion is
triggered. This will generate a core dump which should help investigate
the double close.
As chages to the transaction state are detected by the protocol level
mini-parser, there's no need to fully classify queries inside read-only
transactions. This should be a good performance boost for loads that
heavily use read-only transactions.
If a DCB is closed before a response to the handshake packet is received,
the DCB's session will point to the dummy session. In this case no error
should be written to the DCB.
Now detects some erroneous situations before starting switchover.
Switchover can be activated without specifying current master.
In this case, the cluster master server is selected.
Used the correct value in table_create_alloc and remove unused
parameter. Use the pre-calculated end pointer when looking for events.
Always use the column count of the TABLE_MAP event as all mismatches are
detected earlier.
DATETIME values in old formats should always be 8 bytes long. This is how
MariaDB 10.2 stores them and only DATETIME2 values are stored with a
fractional part.
The parser checks whether the FIRST or AFTER keywords are used and, if
AFTER is used, extracts the relevant column name.
Added a test case that checks that the parsing works and detects the
correct column names.
According to customer reports collecting the statistics has a significant
impact on the performance. As we don't need that information we can just
as well turn off that.
Further, since maxscale-common now links to the sqlite3-library, no
module needs to do that explicitly.
The function used the m_fd member variable before it was updated to point
to the actual file descriptor of the connection. This caused the test to
fail.
Also fixed the nointr_write function to correctly process multiple
consecutive writes.
The CDC connector can be build directly into the core testing library for
testing purposes. This way we remove an unnecessary dependency on a
library. This commit fixes the linkage failure of the cdc_datatypes test.
The follwing statements
SELECT GET_LOCK('lock1',10);
SELECT IS_FREE_LOCK('lock1');
SELECT IS_USED_LOCK('lock1');
SELECT RELEASE_LOCK('lock1');
are now classified as QUERY_TYPE_READ|QUERY_TYPE_WRITE. That will
make cooperative locking work if these functions are used inside
non-read-only transactions and outside transanctions.
If there are several 'users' lines in a rule file, for a particular
user, the rules each matching line will be checked independently
until a rule match is found.
That is, the rules of each 'users' line are treated in an OR-fashion
with respect to each other.
The list of users that is used for authentication shoudl only consist of
users that do not use an explicit authentication plugin. This way
authentication fails before any connections to the backend servers are
done.
If a user has an empty service name, use "mysql" as default.
Authentication data was only updated inside get_pam_user_services() if no service
was found. It was possible that the PAM service changed but the old service
would be used for authenticating, causing a false negative.
Now, the auth data is updated outside the function if authentication fails for
any reason. The new service data is compared to the old and if equal, password
check is not attempted again. This gives a false negative only if user password
has changed after the previous attempt.
Also, fixed some comments.
