repman: add more parameter for failover and switchover

software/repman/buildout.hash.cfg

@@ -18,7 +18,7 @@ md5sum = b38fb4c5cdba2223c84d62ef60513e49
 
 [instance-repman.cfg]
 _update_hash_filename_ = instance-repman.cfg.jinja2.in
-md5sum = 4e9ebd38601658005ad5fe40151e7708
+md5sum = e50536613b175a2e25a041c61adeb59f
 
 [config-toml.in]
 _update_hash_filename_ = templates/config.toml.in
@@ -26,7 +26,7 @@ md5sum = 1efa65af63b14ca2dd1b8c692b2af80e
 
 [config-cluster-toml.in]
 _update_hash_filename_ = templates/cluster-config.toml.in
-md5sum = a0c41a19b5c05752c435b98ad9da935b
+md5sum = 39fb83a9b542b9809948cab98aefa2e7
 
 [nginx.conf.in]
 _update_hash_filename_ = templates/nginx.conf.in

software/repman/instance-repman-input-schema.json

@@ -234,6 +234,95 @@
 							"description": "Listen on IPv6 instead of IPv4.",
 							"type": "boolean",
 							"default": true
+						},
+						"failover-mode": {
+							"title": "Failover mode",
+							"description": "Failover is manual or automatic (default \"automatic\").",
+							"type": "string",
+							"default": "automatic",
+              "enum": [
+                "manual",
+                "automatic"
+              ]
+
+						},
+						"failover-limit": {
+							"title": "Failover amount limit",
+							"description": "Failover is canceld if already failover this number of time (0: unlimited) (default 5).",
+							"type": "integer",
+							"default": 5
+						},
+						"failover-falsepositive-heartbeat": {
+							"title": "Failover check slaves do not receive heartbeat",
+							"description": "Failover checks that slaves do not receive heartbeat (default true).",
+							"type": "boolean",
+							"default": true
+						},
+						"failover-falsepositive-heartbeat-timeout": {
+							"title": "Failover check slaves do not receive heartbeat timeout",
+							"description": "Failover checks that slaves do not receive heartbeat detection timeout  (default 3).",
+							"type": "integer",
+							"default": 3
+						},
+						"failover-falsepositive-ping-counter": {
+							"title": "Failover amount of ping failures",
+							"description": "Failover after this number of ping failures (interval 1s) (default 5).",
+							"type": "integer",
+							"default": 5
+						},
+						"failover-max-slave-delay": {
+							"title": "Failover election ignore slave with replication delay",
+							"description": "Election ignore slave with replication delay over this time in sec (default 30).",
+							"type": "integer",
+							"default": 30
+						},
+						"failover-readonly-state": {
+							"title": "Failover Switchover set slaves as read-only",
+							"description": "Failover Switchover set slaves as read-only (default true).",
+							"type": "boolean",
+							"default": true
+						},
+						"failover-restart-unsafe": {
+							"title": "Failover when cluster down if a slave is start first",
+							"description": "Failover when cluster down if a slave is start first.",
+							"type": "boolean",
+							"default": false
+						},
+						"failover-time-limit": {
+							"title": "Cancel failover if time in sec not passed after previous failover",
+							"description": "Failover is canceled if timer in sec is not passed with previous failover (0: do not wait).",
+							"type": "integer",
+							"default": 0
+						},
+						"switchover-at-equal-gtid": {
+							"title": "Switchover only when slaves are fully in sync",
+							"description": "Switchover only when slaves are fully in sync.",
+							"type": "boolean",
+							"default": false
+						},
+						"switchover-slave-wait-catch": {
+							"title": "Switchover wait for slave to catch with replication",
+							"description": "Switchover wait for slave to catch with replication, not needed in GTID mode but enable to detect possible issues like witing on old master (default true).",
+							"type": "boolean",
+							"default": true
+						},
+						"switchover-wait-kill": {
+							"title": "Switchover wait ms before killing threads on demoted master",
+							"description": "Switchover wait this many milliseconds before killing threads on demoted master (default 5000).",
+							"type": "integer",
+							"default": 5000
+						},
+						"switchover-wait-trx": {
+							"title": "Cancel switchover after timeout seconds if can't aquire FTWRL",
+							"description": "Switchover is cancel after this timeout in second if can't aquire FTWRL (default 10).",
+							"type": "integer",
+							"default": 10
+						},
+						"switchover-wait-write-query": {
+							"title": "Cancel switchover after timeout if a write query is running",
+							"description": "Switchover is canceled if a write query is running for this time (default 10).",
+							"type": "integer",
+							"default": 10
 						}
 					},
 					"type": "object"

software/repman/instance-repman.cfg.jinja2.in

@@ -122,6 +122,22 @@ db-disk-iops = {{ parameter_dict.get("db-disk-iops", 300) }}
 db-memory = {{ parameter_dict.get("db-memory", 256) }}
 db-memory-shared-pct = {{ parameter_dict.get("db-memory-shared-pct", "threads:16,innodb:60,myisam:10,aria:10,rocksdb:1,tokudb:1,s3:1,archive:1,querycache:0") }}
 db-memory-threaded-pct = {{ parameter_dict.get("db-memory-threaded-pct", "tmp:70,join:20,sort:10") }}
+# failover
+failover-mode = {{ parameter_dict.get('failover-mode', 'automatic') }}
+failover-limit = {{ parameter_dict.get('failover-limit', 5) }}
+failover-falsepositive-heartbeat = {{ parameter_dict.get('failover-falsepositive-heartbeat', True) }}
+failover-falsepositive-heartbeat-timeout = {{ parameter_dict.get('failover-falsepositive-heartbeat-timeout', 3) }}
+failover-falsepositive-ping-counter = {{ parameter_dict.get('failover-falsepositive-ping-counter', 5) }}
+failover-max-slave-delay = {{ parameter_dict.get('failover-max-slave-delay', 30) }}
+failover-readonly-state = {{ parameter_dict.get('failover-readonly-state', True) }}
+failover-restart-unsafe = {{ parameter_dict.get('failover-restart-unsafe', False) }}
+failover-time-limit = {{ parameter_dict.get('failover-time-limit', 0) }}
+#switchover
+switchover-at-equal-gtid = {{ parameter_dict.get('switchover-at-equal-gtid', False) }}
+switchover-slave-wait-catch = {{ parameter_dict.get('switchover-slave-wait-catch', True) }}
+switchover-wait-kill = {{ parameter_dict.get('switchover-wait-kill', 5000) }}
+switchover-wait-trx = {{ parameter_dict.get('switchover-wait-trx', 10) }}
+switchover-wait-write-query = {{ parameter_dict.get('switchover-wait-write-query', 10) }}
 
 [{{ 'config-' ~ name }}]
 recipe = slapos.recipe.template:jinja2

software/repman/templates/cluster-config.toml.in

+{% macro setbool(value) -%}
+{% if value in ['true', 'True'] %} true {% else %} false {% endif -%}
+{% endmacro -%}
 [{{ parameter_dict['name'] }}]
 title = "{{ parameter_dict['name'] }}"
 monitoring-save-config = true
@@ -43,4 +46,22 @@ prov-db-memory = {{ parameter_dict['db-memory'] }}
 prov-db-memory-shared-pct = "{{ parameter_dict['db-memory-shared-pct'] }}"
 prov-db-memory-threaded-pct = "{{ parameter_dict['db-memory-threaded-pct'] }}"
 
-test-inject-traffic = true
\ No newline at end of file
+test-inject-traffic = true
+# failover
+failover-mode = "{{ parameter_dict['failover-mode'] }}"
+failover-limit = {{ parameter_dict['failover-limit'] }}
+failover-falsepositive-heartbeat = {{ setbool(parameter_dict['failover-falsepositive-heartbeat']) }}
+failover-falsepositive-heartbeat-timeout = {{ parameter_dict['failover-falsepositive-heartbeat-timeout'] }}
+failover-falsepositive-ping-counter = {{ parameter_dict['failover-falsepositive-ping-counter'] }}
+failover-max-slave-delay = {{ parameter_dict['failover-max-slave-delay'] }}
+failover-readonly-state = {{ setbool(parameter_dict['failover-readonly-state']) }}
+failover-restart-unsafe = {{ setbool(parameter_dict['failover-restart-unsafe']) }}
+failover-time-limit = {{ parameter_dict['failover-time-limit'] }}
+#switchover
+switchover-at-equal-gtid = {{ setbool(parameter_dict['switchover-at-equal-gtid']) }}
+switchover-slave-wait-catch = {{ setbool(parameter_dict['switchover-slave-wait-catch']) }}
+switchover-wait-kill = {{ parameter_dict['switchover-wait-kill'] }}
+switchover-wait-trx = {{ parameter_dict['switchover-wait-trx'] }}
+switchover-wait-write-query = {{ parameter_dict['switchover-wait-write-query'] }}
+
+