revert(infra): remove max_connect_errors=100000 from dbdeployer entrypoints

The real fix for the blocked host errors was granting SUPER/REPLICATION
SLAVE to the monitor user. Inflating max_connect_errors just masked the
problem. Let MySQL use its default (100).

test/infra/infra-dbdeployer-mariadb10/docker/entrypoint.sh

@@ -23,7 +23,6 @@ dbdeployer deploy replication "${MYSQL_VERSION}" \
     --base-port=3305 \
     -c log-slave-updates \
     -c max_connections=500 \
-    -c max_connect_errors=100000 \
     -c max_binlog_size=100M \
     -c plugin_load_add=ha_blackhole
 

test/infra/infra-dbdeployer-mysql57-binlog/docker/entrypoint.sh

@@ -25,7 +25,6 @@ dbdeployer deploy replication "${MYSQL_VERSION}" \
     -c log-slave-updates \
     -c binlog_format=ROW \
     -c max_connections=500 \
-    -c max_connect_errors=100000 \
     -c innodb_buffer_pool_size=128M \
     -c innodb_log_file_size=32M \
     -c innodb_flush_log_at_trx_commit=2 \

test/infra/infra-dbdeployer-mysql57/docker/entrypoint.sh

@@ -25,7 +25,6 @@ dbdeployer deploy replication "${MYSQL_VERSION}" \
     -c log-slave-updates \
     -c binlog_format=ROW \
     -c max_connections=500 \
-    -c max_connect_errors=100000 \
     -c innodb_buffer_pool_size=128M \
     -c innodb_log_file_size=32M \
     -c innodb_flush_log_at_trx_commit=2 \

test/infra/infra-dbdeployer-mysql84-binlog/docker/entrypoint.sh

@@ -26,7 +26,6 @@ dbdeployer deploy replication "${MYSQL_VERSION}" \
     -c log-replica-updates \
     -c binlog_format=ROW \
     -c max_connections=500 \
-    -c max_connect_errors=100000 \
     -c local_infile=ON \
     -c innodb_buffer_pool_size=128M \
     -c innodb_redo_log_capacity=64M \

test/infra/infra-dbdeployer-mysql84/docker/entrypoint.sh

@@ -26,7 +26,6 @@ dbdeployer deploy replication "${MYSQL_VERSION}" \
     -c log-replica-updates \
     -c binlog_format=ROW \
     -c max_connections=500 \
-    -c max_connect_errors=100000 \
     -c local_infile=ON \
     -c innodb_buffer_pool_size=128M \
     -c innodb_redo_log_capacity=64M \