query_classifier.h: added QUERY_TYPE_CREATE_TMP_TABLE and QUERY_TYPE_READ_TMP_TABLE for use of temporary table support.
hashtable.c:Added variant of hashtable which is 'flat', that is, stored to existing memory instead of allocating memory as a part of the call. Existing function declarations don't change but added hashtable_alloc_flat for the purpose. Both hashtable_alloc and hashtable_alloc_flat now call the real allocation function, hashtable_alloc_real. hashtable_free only frees memory which is allocated in hashtable_alloc_real.
hashtable.h: added a flag to HASHTABLE struct to indicate whether hashtable owns its memory or not.
readwritesplit.h: Added RSES_PROP_TYPE_TMPTABLES property type to be used for keeping the hashtable for tablenames.
readwritesplit.c: Added comments about temporary table support implementation.
query_classifier.h: added parsing information structure to query classifier away from buffer.h.
buffer.c:introduced a buffer object which includes object pointer and a clean-up call-back function. Buffer objects form a list which is cleaned up by the last referrer of the buffer, as a part of gwbuf_free. Buffer object list is protected by a spinlock gwbuf_lock.
Also added identifier type, bufobj_id_t which is enumerated type and currently includes one value only, GWBUF_PARSING_INFO. Added also a bitfield for information about the buffer. It currently has one type only, GWBUF_INFO_PARSED indicating that buffer content is parsed and there is buffer object of type GWBUF_PARSING_INFO.
skygw_utils.cc:replace_literal:changed regexec matching to case insensitive because user-defined literals are sometimes converted to upper-case ones.
skygw_utils.cc: removed replace_str and implemented replace_literal which replaces user-provided literals in query with predefined string "?". Replacing is done one by one, so it is suboptimal ipmlementation since all literals could be passed to replacement function in one call and processed all before returning.
query_classifier.cc: Now query can be parsed outside query_classifier_get_type by calling function parse_query. It creates parsing_info_t struct which is then added to the GWBUF which also includes the query. Parsing information follows the buffered query and it is freed at the same time with query buffer, in gwbuf_free.
buffer.c: additions of parsing information to gwbuf struct.
modutil.c: added function which returns query from GWBUF in plain text string.
readwritesplit.c:routeQuery now only calls query_classifier_get_type to get the query type instead of extracting plain text query from the GWBUF buffer.
* ignore typical backup files created by common editors
* move general ignore rules like "*.o" or "depend.mk" to top level gitignore
* ignore executables and test directories in target dir gitignore
as these are local and there's no general catch-all pattern for them
All prepare commands are executed in every backend server currently connected.
All executes are routed to master. If stmt type was recorded in prepare phase in rwsplit router, read-only stmts could be routed to slaves.
COM_STMT_PREPARE gets arbitrary number of response packets from backend database. Since statements are prepared in every backend server and only one multi-packet response can be replied to client, redundant multi-packet responses are discarded. This is done in router. Mechanisms from session command handling are utilized with little changes: router must identify when response consists of multiple packets so that it knows to calculate the number of packets in response and that it is able to discard correct number of packets.
Information to the reply-handling router is provided by backend protocol, which includes a ordered list of commands of commands sent to protocol-owning backend server. A command is stored to protocol struct in mysql_backend.c:gw_MySQLWrite_backend if the statement buffer's type has GWBUF_TYPE_SINGLE_STMT set in mysql_client.c:route_by_statement. GWBUF_TYPE_SINGLE_STATEMENT indicates that there is single statement in the buffer, as opposite to Read Connection router, which accepts streaming input from client.
Added a lot of logging to error, trace and message logs which should help the user to handle errors which can't be automatically resolved, like attempt to use nonexisting database.
query_classifier.cc use of uninitialized value in skygw_stmt_causes_implicit_commit
config.c crashed if module load failed, use of unitialized value
load_utils.c pretty-printed error
service.c use of uninitialized value in service_add_qualified_param
modules.h function prototype
readwritesplit.c memory leaks
1. in query_classifier.cc autocommit_enabled, and transaction_active variables maintained their values across different sessions. Now those values are stored in each router_client_ses object.
2. As a part of implementation of MAX-95 session variables were added to BACKEND struct which is shared with all sessions using the SERVICE which the particular BACKEND serves. Now each router_client_ses object has a backend reference struct which includes pointer to BACKEND, DCB and to session command cursor.
Added test - set_autocommit_disabled.sql, test_after_autocommit_disabled.sql - to check that session variable is discarded when session where it belongs terminates.
All tests are run by executing 'make testall' in root directory. As a result all directories which contain tests will be entered and tests executed.
After tests, each directory have a test log including the last run's logs only.
Created a global test log, which is specified in test.inc. Test logs from subdirectories are concatenated to this file.
Increased skygw_query_type_t to 16 bits, and corrected the way how those bit fields are checked.
Added tests for cases where autocommit is disabled and corrected old tests.
Every test/makefile have the following targets:
cleantests - clean local and subdirectories' tests
buildtests - build all local and subdirectories' tests
runtests - run all local tests
testall - clean, build and run local and subdirectories' tests
Tests for directory random_dir are always in its subdirectory, in this case in random_dir/test . If random_dir has subdirectories with tests, random_dir/child_dir, for example, tests of child_dir can be started from random_dir/test/makefile where make -C child_dir/test <test target> is called.
See MAXSCALE_HOME/test/README for further information.
Bug was identified by test, but it couldn't be validated. Using RW Split router currently may lead to situation where after setting system variable there is a backend server which doesn't have the value set. This is an incomplete feature at the moment rather than a bug.
