It should be possible to START SLAVE and STOP SLAVE irrespective
of which Galera node updates are mode to.
That will be the case if @@log_slave_updates is on and each node
in the Galera cluster have the same server id. Otherwise it will
fail with the current incarnation of BLR.
If
* BLR replicates from a node in a Galera cluster and
* writes are made to all nodes in that cluster,
then
* if a slave to BLR is stopped when it has received an event
originating in a node different than the one BLR is replicating
from
the subsequent (re)starting of the slave will fail because BLR looks
for the last event from a file whose path contains the server id of
the node where the event originates, although it should look for it
in the file whose path contains the server id of the node from which
BLR replicates.
The test adds events to the backends, causing them to go out of sync. It then
uses the reset-replication command to fix the situation and checks replication
is working.
The test adds a scheduled server event, the does failover, rejoin and
switchover and checks that event is manipulated correctly. Also includes
a change to the monitor to fix an invalid ALTER EVENT query when the event
definer has wildcard host.
The behavior of mariadbmon was changed so that it better understands
slaves attempting to replicate. Rewrote the test to accommodate the change
in behavior and take the opportunity to use newer code.
Because of monitor changes, the test had wrong assumptions.
Renamed the test and updated it to use MaxCtrl for some queries.
Also, changed the type of the cycle container in the monitor to an
ordered map so that results are predictable.
After a session command fails on all connected slaves but succeeds on the
master, the servers that failed are discarded from the pool of valid
servers. If there are available servers, they will be taken into use when
the next read query is performed.
When a query is routed to an unconnected slave, it is taken into use and
the session command history is replayed. If the execution of the history
failed and there were more session commands to be executed, any queued
queries would get lost. This is due to a missing check that would detect
the fact that the server has been discarded.
The test now sets a two minute timeout for all larger operations. This
prevents excessive waiting when the test is executed.
Removed the row count output to stdout to prevent excessive logging when
the terminal contents are written to a file.
Also renamed the file to match the test case name. This should remove the
need to find the source file to test name mapping from the CMakeLists.txt.
