Removed old router property code as it is no longer needed when
SessionCommand class used by the Backend class is taken into use.
Removed unnecessary code that is implemented as a part of the Backend
class.
Changed functions to return references to Backends instead of handling raw
DCBs. This introduces a few cases where the code returns a reference when
no reference is actually available. These cases are solved by having an
empty static shared_ptr that is returned in these cases. This is done to
silence any compiler warnings that returning references to local variables
would bring as these should never happen if the code is functioning
properly.
This is the first step to taking the Backend class into use. It is now
used in rwsplit_select_backends.cc and readwritesplit.hh. The module is
not yet functional and doesn't even compile.
Added some helper functions to the Backend class to get easier access to
the server referenced by the SERVER_REF and to check the state of the
backend.
Creating duplicate connections using the Backend class allows the
connections and their handling to be tested at the same time that the old
system is in place. This should make it somewhat easier to grasp what
changes and where when the new implementation is taken into use.
The session commands are now duplicated as SessionCommand objects This
allows for an easier migration from the old session command implementation
to the new one.
The states are now internal to the Backend class. This simplifies the use
of the class by moving the burder of state tracking to the class
itself.
Refactored the way the schemarouter uses the Backend class.
Also fixed a memory leak in the schemarouter when `ignore_databases_regex`
was used..
Using the same implementation of the Backend and SessionCommand classes in
both schemarouter and readwritesplit will prevent duplication of code.
This commit only splits the generic parts of the class to a Backend class
which the schemarouter then extends. The session commands for both routers
are similar so they require no special handling.
The default database can now be manipulated with a set of functions
exposed by the maxscale/protocol/mysql.h header. This removes the need to
handle the structures themselves in the modules and is a step towards
moving the dcb->data contents inside the session.
Reorganized the main source file of readwritesplit. The internal functions
are first followed by the API entry points. The actual module definition
is the last declared object in the file.
The temporary table detection and handling now uses C++ containers to
store the set of temporary tables. The detection also uses the new query
classifier field info API to detect which tables and databases are
targeted.
Cleaned up the readwritesplit main header. The structs were named to their
typedef names so that no code changes are required. The structs should be
renamed if/when they are converted to proper C++ classes.
The tee filter was missing the match/exclude functionality that was in the
previous implementation. The filter was upgraded to use PCRE2 regular
expressions instead of POSIX regular expressions.
Documented that the match/exclude patterns should use PCRE2 syntax.
The function should actually be in include/maxscale/protocol/mysql.h
and the implementation in MySQLCommon, but the monitors do not link
to that yet.
All MySQL related should be moved to MySQLCommon and the core
refactored so that no MySQL knowledge is needed there.
With this change, it is no longer possible that the server version
is deallocated at the very moment it is read. There is still a race,
but it's mostly harmless.
The HintParser wrongly ignored linebreaks, causing parsing faults
e.g. parsing too far or accepting invalid comments. Now, the parser
detects a line break and terminates comments unless they started with
'/*'. Also, fixed a memory leak when parsing parameter-value-combinations.
Whenever a server which is a slave of an external master is detected, it
will be assigned the slave status. This will allow the status to be used
the way it was intended to be used.
The modutil_get_SQL()-function allocates storage, while
modutil_extract_SQL() does not. The strings given by the latter
are not 0-terminated so require a length limit when matched using
regexec().
This commit changes the used function in those cases where the
sql-string is not modified nor is the pointer saved for later use.
MXS-1266:
blr_slave_gtid_request() the file in slave registration request is
checked whether it exists or not and GTID file info could e set instead.
Also addded:
blr_handle_simple_select_stmt(): only mariadb10_gtid_domain is checked
for SELECT @@GLOBAL.gtid_domain_id slave request
blr_start_master_registration(): only router->mariadb10_compat is
checked before sending SELECT @@GLOBAL.gtid_domain_id to master
blr_file_create calls mxs_mkdir_all()
MASTER_USE_GTID=Slave_pos is now handled by CHANGE MASTER TO
If mariadb10_master_gtid is On
MASTER_LOG_FILE is no longer required, only MASTER_USE_GTID=Slave_pos
Slave_pos must be set before to empty value or request value:
set @@global.gtid_slave_pos = '0-10116-194';
MXS-1067. The CCRFilter will now look at the hints in a write-type
query. The parameter-value combinations "ccr=match" and "ccr=ignore"
are recognized and will override any regex match and ignore settings.
For the hints to work, the HintFilter needs to be present in the filter
chain before the CCRFilter.
Also ran astyle and updated documentation.
The module commands can now produce JSON formatted output which is passed
to the caller. The output should conform to the JSON API as closely as
possible.
Currently, the REST API wraps all JSON produced by module commands inside
a meta-object of the following type:
{
"meta": <output of module command>
}
This allows the output to be JSON API conformant without modifying the
modules and allows incremental updates to code.
The MySQL monitor replication_heartbeat table now uses the default storage
engine of the database when creating the table. Most of the time the
default is InnoDB which makes the table crash-safe.
The addition of field types and lengths wasn't added to the avrorouter
ALTER TABLE handler. This caused crashes when an alter table was done and
new rows were inserted afterwards.
