Problem:- Gtid are not transferred in Galera Cluster.
Solution:- We need to transfer gtid in the case on either when cluster is
slave/master in async replication. In normal Gtid replication gtid are generated on
recieving node itself and it is always on sync with other nodes. Because galera keeps
node in sync , So all nodes get same no of event groups. So the issue arises when
say galera is slave in async replication.
A
| (Async replication)
D <-> E <-> F {Galera replication}
So what should happen is that all node should apply the master gtid but this does
node happen, becuase node E, F does not recieve gtid from D in write set , So what E(or F)
does is that it applies wsrep_gtid_domain_id, D server-id , E gtid next seq no. This
generated gtid does not always work when say A has different domain id.
So In this commit, on galera node when we see that this event is recieved from master
we simply write Gtid_Log_Event in write_set and send it to other nodes.
A suggestion to make role propagation simpler from serg@mariadb.org.
Instead of gathering the leaf roles in an array, which for very wide
graphs could potentially mean a big part of the whole roles schema, keep
the previous logic. When finally merging a role, set its counter
to something positive.
This will effectively mean that a role has been merged, thus a random pass
through roles hash that touches a previously merged role won't cause the problem
described in MDEV-12366 any more, as propagate_role_grants_action will stop
attempting to merge from that role.
In the function make_sortkey a tmp buffer was defined and in the absence of
param->tmp_buffer, tmp buffer used the sort_keys buffer. sort_keys buffer
has a length defined in sort_field->length, while param->tmp_buffer is
stored in param->rec_length. Make sure to use the appropriate length
based on which buffer we are using otherwise we'll overflow.
Also added a type cast to size_t during the calculation of the sort keys
buffer size to avoid an oveflow if the buffer size exceeds 32 bits.
galera_events test shows a regression with the original fix for MW-416
Reason was that Events::drop_event() can be called also from inside event
execution, and there we have a speacial treatment for event, which executes
"DROP EVENT" statement, and runs TOI replication inside the event processing body.
This resulted in executing WSREP_TO_ISOLATION two times for such DROP EVENT statement.
Fix is to call WSREP_TO_ISOLATION_BEGIN only in Events::drop_event()
Changed return code for replicatio error to TRUE.
This is aligned with native mysql convention to return TRUE (defined to 1) or FALSE (defined to 0) from a bool function.
This is wrong, but follows the mysql conventiosn, at least...
find_type_or_exit() client helper did exit(1) on error, exit(1) moved to
clients.
mysql_read_default_options() did exit(1) on error, error is passed through and
handled now.
my_str_malloc_default() did exit(1) on error, replaced my_str_ allocator
functions with normal my_malloc()/my_realloc()/my_free().
sql_connect.cc did many exit(1) on hash initialisation failure. Removed error
check since my_hash_init() never fails.
my_malloc() did exit(1) on error. Replaced with abort().
my_load_defaults() did exit(1) on error, replaced with return 2.
my_load_defaults() still does exit(0) when invoked with --print-defaults.
Whenever we call merge_role_privileges on a role, we make use of
the role->counter variable to check if all it's children have had their
privileges merged. Only if all children have had their privileges merged,
do we update the privileges on parent. This is done to prevent extra work.
The same idea is employed during flush privileges. You only begin merging
from "leaf" roles. The recursive calls will merge their parents at some point.
A problem arises when we try to "re-merge" a parent. Take the following graph:
{noformat}
A (0) ---- C (2) ---- D (2) ---- USER
/ /
B (0) ----/ /
/
E (0) --------------/
{noformat}
In parentheses we have the "counter" value right before we start to iterate
through the roles hash and propagate values. It represents the number of roles
granted to the current role. The order in which we iterate through the roles
hash is alphabetical.
* First merge A, which leads to decreasing the counter for C to 1. Since C is
not 0, we don't proceed with merging into C.
* Second we merge B, which leads to decreasing the counter for C to 0. Now
we proceed with merging into C. This leads to reducing the counter for D to 1
as part of C merge process.
* Third as we iterate through the hash, we see that C has counter 0, thus we
start the merge process *again*. This leads to reducing the counter for
D to 0! We then attempt to merge D.
* Fourth we start merging E. When E sees D as it's parent (according to the code)
it attempts to reduce D's counter, which leads to overflow. Now D's counter is
a very large number, thus E's privileges are not forwarded to D yet.
To correct this behavior we must make sure to only start merging from initial
leaf nodes.
When granting a role to another role, DB privileges get propagated. If
the grantee had no previous DB privileges, an extra ACL_DB entry is created to
house those "indirectly received" privileges. If, afterwards, DB
privileges are granted to the grantee directly, we must make sure to not
create a duplicate ACL_DB entry.
Moved TOI replication to happen after ACL checking for commands:
SQLCOM_CREATE_EVENT
SQLCOM_ALTER_EVENT
SQLCOM_DROP_EVENT
SQLCOM_CREATE_VIEW
SQLCOM_CREATE_TRIGGER
SQLCOM_DROP_TRIGGER
SQLCOM_INSTALL_PLUGIN
SQLCOM_UNINSTALL_PLUGIN
This was missing bug fix from MySQL wsrep i.e. Galera.
Problem was that if stored procedure declares a handler that
catches deadlock error, then the error may have been
cleared in method sp_rcontext::handle_sql_condition().
Use wsrep_conflict_state correctly to determine is the
error already sent to client.
Add test case for both this bug and MDEV-12837: WSREP: BF
lock wait long. Test requires both fixes to pass.
This is 10.1 version where no merge error exists.
wsrep_on_check
New check function. Galera can't be enabled
if innodb-lock-schedule-algorithm=VATS.
innobase_kill_query
In Galera async kill we could own lock mutex.
innobase_init
If Variance-Aware-Transaction-Sheduling Algorithm (VATS) is
used on Galera we refuse to start InnoDB.
Changed innodb-lock-schedule-algorithm as read-only parameter
as it was designed to be.
lock_rec_other_has_expl_req,
lock_rec_other_has_conflicting,
lock_rec_lock_slow
lock_table_other_has_incompatible
lock_rec_insert_check_and_lock
Change pointer to conflicting lock to normal pointer as this
pointer contents could be changed later.
This reverts commit 7603463a46.
The commit itself is fine, however when disabling volatile, compiler
optimizations mess up our double results due to precision differences.
Revert the patch till a proper solution is found.
This was added in c796415943 but would hurt all other compilers
because of Visual Studio. Hopefully this has been fixed now.
Signed-off-by: Daniel Black <daniel@linux.vnet.ibm.com>
For BIT field null_bit is not set to 0 even for a field defined as NOT NULL.
So now in the function TABLE::create_key_part_by_field, if the bit field is not nullable
then the null_bit is explicitly set to 0
Issue:
------
VALUES doesn't have a type() function and is considered a
Item_field.
Solution for 5.7:
-----------------
Add a new type() function for Item_values_insert.
On 8.0 and trunk it was fixed by Mithun's Bug#19601973.
Solution for 5.6:
-----------------
Additionally Bug#17458914 is backported.
This will address the problem of using VALUES() in
INSERT ... ON DUPLICATE KEY UPDATE. Create a field object
only if it is in the UPDATE clause, else return a NULL
item.
This will also address the problems mentioned in
Bug#14789787 and Bug#16756402.
Solution for 5.5:
-----------------
As mentioned above Bug#17458914 is backported.
Additionally Bug#14786324 is also backported.
When VALUES() is detected outside its meaningful place,
it should be treated as NULL and is thus replaced with a
Field_null object, with the same name as the original
field.
Fields with type NULL are generally not handled well inside
the server (e.g Innodb will not accept them and it is
impossible to create them in regular tables). So create a
new const NULL item instead.
With combination of --log-bin and Galera the server may crash
reporting two characteristic stacks:
/usr/sbin/mysqld(_ZN13MYSQL_BIN_LOG13mark_xid_doneEmb+0xc7)[0x7f182a8e2cb7]
/usr/sbin/mysqld(binlog_background_thread+0x2b5)[0x7f182a8e3275]
or
/usr/sbin/mysqld(_ZN13MYSQL_BIN_LOG21do_checkpoint_requestEm+0x9d)[0x7ff395b2dafd]
/usr/sbin/mysqld(_ZN13MYSQL_BIN_LOG20checkpoint_and_purgeEm+0x11)[0x7ff395b2db91]
/usr/sbin/mysqld(_ZN13MYSQL_BIN_LOG16rotate_and_purgeEb+0xc2)[0x7ff395b300b2]
The reason of the failure appears to be non-matching decrements for
`xid_count_per_binlog::xid_count`
which can occur when a transaction is executed having its connection issued
`SET @@sql_log_bin=0`. In such case the xid count is not incremented but
its decrements still runs to turn `binlog_xid_count_list` into improper state
which the following FLUSH BINARY LOGS exposes through the crash.
*Note_1*: the regression test reuses an existing galera.sql_log_bin
which does not run stably (even in its base form) by mtr with --log-bin.
*Note_2*: 10.0-galera branch is free of this issue having missed MDEV-7205
fixes.
As reported in MDEV-11969 "there's no way to ditch knowledge" about some
domain that is no longer updated on a server. Besides being of annoyance to
clutter output in DBA console stale domains can prevent the slave
to connect the master as MDEV-12012 witnesses.
What domain is obsolete must be evaluated by the user (DBA) according
to whether the domain info is still relevant and will the domain ever
receive any update.
This patch introduces a method to discard obsolete gtid domains from
the server binlog state. The removal requires no event group from such
domain present in existing binlog files though. If there are any the
containing logs must be first PURGEd in order for
FLUSH BINARY LOGS DELETE_DOMAIN_ID=(list-of-domains)
succeed. Otherwise the command returns an error.
The list of obsolete domains can be computed through
intersecting two sets - the earliest (first) binlog's Gtid_list
and the current value of @@global.gtid_binlog_state - and extracting
the domain id components from the intersection list items.
The new DELETE_DOMAIN_ID featured FLUSH continues to rotate binlog
omitting the deleted domains from the active binlog file's Gtid_list.
Notice though when the command is ineffective - that none of requested to delete
domain exists in the binlog state - rotation does not occur.
Obsolete domain deletion is not harmful for connected slaves as long
as master side binlog files *purge* is synchronized with FLUSH-DELETE_DOMAIN_ID.
The slaves must have the last event from purged files processed as usual,
in order not to bump later into requesting a gtid from a file which
was already gone.
While the command is not replicated (as ordinary FLUSH BINLOG LOGS is)
slaves, even though having extra domains, won't suffer from reconnection errors
thanks to master-slave gtid connection protocol allowing the master
to be ignorant about a gtid domain.
Should at failover such slave to be promoted into master role it may run
the ex-master's
FLUSH BINARY LOGS DELETE_DOMAIN_ID=(list-of-domains)
to clean its own binlog state.
NOTES.
suite/perfschema/r/start_server_low_digest.result
is re-recorded as consequence of internal parser codes changes.
uses alias in HAVING when sql_mode = 'ONLY_FULL_GROUP_BY'
This patch corrects the patch for bug#18739: non-standard
HAVING extension was allowed in strict ANSI sql mode
added in 2006 by commit 4b7c4cd27f.
As a result of incompleteness of the fix in the above commit
if a query with GROUP BY contained an aggregate function with an
alias and this alias was used in the HAVING clause of the query
the server reported an error when sql_mode was set to
'ONLY_FULL_GROUP_BY'.
Make differentiation between pullout for merge and pulout of outer field during exists2in transformation.
In last case the field was outer and so we can safely start from name resolution context of the SELECT where it was pulled.
Old behavior lead to inconsistence between list of tables and outer name resolution context (which skips one SELECT for merge purposes) which creates problem vor name resolution.
* created tests focusing in multi-master conflicts during cascading foreign key
processing
* in row0upd.cc, calling wsrep_row_ups_check_foreign_constraints only when
running in cluster
* in row0ins.cc fixed regression from MW-369, which caused crash with MW-402.test
* changed thd_binlog_format to return configured binlog format in wsrep execution,
regardless of binlogging setting (i.e. with or without binlogging)
* thd_binlog_format is used in innobase::write_row(), and would return confusing
result there when log_bin==OFF
