Reconnect to mgm server in connect_clients thread if m_mgm_handle is disconnected.
ndb/include/mgmapi/mgmapi.h:
Add prototype for ndb_mgm_is_connected
ndb/src/common/transporter/TransporterRegistry.cpp:
start_clients_thread:
- don't get port if we don't have to
- Reconnect to mgm server if m_mgm_handle is disconnected
ndb/src/mgmapi/mgmapi.cpp:
add ndb_mgm_is_connected
Can't repeat information_schema test failure
Fix for lowercase_table3 only
mysql-test/r/lowercase_table3.result:
Fix for bug#8156:information_schema and lowercase_table3 fail on OSX
mysql-test/t/lowercase_table3.test:
Fix for bug#8156:information_schema and lowercase_table3 fail on OSX
sql/ha_innodb.cc:
Fix for bug#8156:information_schema and lowercase_table3 fail on OSX
Make sure that the node itself is set in the m_connected_nodes when connected
ndb/src/kernel/vm/Configuration.cpp:
make sure we destroy the iterator before reusing the variable
ndb/src/mgmsrv/MgmtSrvr.cpp:
make sure that the "conections are up to date" when exiting from MgmtSrvr::transporter_connect()
ndb/src/ndbapi/ClusterMgr.cpp:
Make sure that the node itself is set in the m_connected_nodes when connected
Fix a glitch reported by Guilhem: release the memory allocated for innodb_internal_... if innobase_init() fails; cleanups: in ha_innodb, you should use the InnoDB formatting style for the source code
sql/ha_innodb.cc:
Fix a glitch reported by Guilhem: release the memory allocated for innodb_internal_... if innobase_init() fails; cleanups: in ha_innodb, you should use the InnoDB formatting style for the source code
suggested in review by Heikki.
innobase/include/data0type.h:
Improve comment of dtype_struct.
innobase/include/data0type.ic:
Rename innobase_get_mb_cset() to innobase_get_cset_width().
dtype_get_fixed_size(): Add an extra integrity check for
type->prtype, type->mbminlen and type->mbmaxlen.
sql/ha_innodb.cc:
Rename innobase_get_mb_cset() to innobase_get_cset_width()
use m_cofigMutex when configuration is accessed
use c++ iteration interface everywhere in MgmtSrvr code
use NDB_INVALID_SOCKET everywhere in Services.cpp
ndb/src/common/transporter/TransporterRegistry.cpp:
removed warning
ndb/src/mgmapi/mgmapi.cpp:
mgmapi to return connected port and interface _only_ if it is actually connected
ndb/src/mgmsrv/MgmtSrvr.cpp:
use m_cofigMutex when configuration is accessed
use c++ iteration interface everywhere
ndb/src/mgmsrv/Services.cpp:
use NDB_INVALID_SOCKET everywhere
connect_client thread will work more than just first-time.
ndb/src/common/transporter/Transporter.cpp:
m_r_port is now m_s_port a SIGNED port. <0 is dynamic
ndb/src/common/transporter/Transporter.hpp:
m_r_port is now m_s_port, a SIGNED port number (<0 is dynamic).
We give the socket client a positive port though.
ndb/src/common/transporter/TransporterRegistry.cpp:
Transporter now holds SIGNED port number (<0 dynamic)
Fix memory corruption bug reported by Georg Richter: if one created a table whose primary key contained at least 2 column prefixes, that could corrupt the memory heap of prebuilt_t
innobase/row/row0mysql.c:
Fix memory corruption bug reported by Georg Richter: if one created a table whose primary key contained at least 2 column prefixes, that could corrupt the memory heap of prebuilt_t
into mysql.com:/usr/local/home/marty/MySQL/mysql-5.0-ndb
BitKeeper/etc/logging_ok:
auto-union
BitKeeper/triggers/post-commit:
Auto merged
sql/item.cc:
Auto merged
sql/log.cc:
Auto merged
BitKeeper/etc/logging_ok:
auto-union
BitKeeper/triggers/post-commit:
Auto merged
include/my_pthread.h:
Auto merged
mysql-test/r/select.result:
Auto merged
mysql-test/t/select.test:
Auto merged
sql/item.cc:
Auto merged
sql/log.cc:
Auto merged
sql/table.cc:
SCCS merged
into rurik.mysql.com:/home/igor/dev/mysql-4.1-0
BitKeeper/etc/logging_ok:
auto-union
extra/replace.c:
Auto merged
mysql-test/r/select.result:
Auto merged
mysql-test/t/select.test:
Auto merged
sql/table.cc:
Auto merged
Added a test case for bug #7520.
table.cc:
Fixed bug #7520.
The bug was caused by a wrong calculation of the field max_key_length for
a TABLE structure when there was an index on a blob field.
sql/table.cc:
Fixed bug #7520.
The bug was caused by a wrong calculation of the field max_key_length for
a TABLE structure when there was an index on a blob field.
mysql-test/t/select.test:
Added a test case for bug #7520.
mysql-test/r/select.result:
Added a test case for bug #7520.
into mysql.com:/home/jonas/src/mysql-5.0
BitKeeper/etc/logging_ok:
auto-union
BitKeeper/triggers/post-commit:
Auto merged
ndb/test/run-test/daily-devel-tests.txt:
Auto merged
sql/item.cc:
Auto merged
sql/log.cc:
Auto merged
- when connecting to a mgm server as a transporter, create a NdbMgmHandle
- over this mgm handle, report the dynamic ports
- then turn it into a transporter
- this will re-report dynamic ports to mgmds when they restart (as we'll have to
set up our transporter again). This will also report it to all mgmds (as we'll
have transporters to all of them).
ndb/include/mgmapi/mgmapi.h:
Add ndb_mgm_convert_to_transporter
- converts to a transporter connect
- destroys the handle (without disconnecting)
- returns socket
ndb/include/transporter/TransporterRegistry.hpp:
Add prototype for connect_ndb_mgmd
ndb/include/util/SocketClient.hpp:
Remove connect_without_auth
Add get_port() and get_server_name()
ndb/src/common/transporter/Transporter.cpp:
use TransporterRegistry::connect_ndb_mgmd() to make the connection if isMgmConnection
ndb/src/common/transporter/TransporterRegistry.cpp:
Impliment TransporterRegistry::connect_ndb_mgmd
- takes a SocketClient and constructs a connectstring.
- uses this connect string to make a NdbMgmHandle
- send dynamic ports to this mgm server
- transform into a transporter connect
- return socket
ndb/src/common/util/SocketClient.cpp:
Remove connect_without_auth
ndb/src/kernel/main.cpp:
Don't relay dynamic ports. We now do this on transporter connect
ndb/src/mgmapi/mgmapi.cpp:
Impliment ndb_mgm_convert_to_transporter
- converts the mgm connection into a transporter connection
- destroys the handle
- returns the socket that should now be used as a transporter
