alwayson監控

最近大致本身寫了點alwayson相關的監控，是經過存儲過程的方法，作個筆記以下：

--alwayson啓用狀態
declare @isenabled sql_variant
SELECT @isenabled=SERVERPROPERTY ( 'IsHadrEnabled' ) 
if @isenabled=1
begin
print 'is enabled（alwayson啓用狀態）'
end
---------監控集羣的狀態
if exists(SELECT * FROM sys.dm_hadr_cluster_members where member_state <>1)
begin
    print 'alwayson集羣有拖機機器，請查看'
end
------------監控副本同步狀態
if exists(select * from sys.dm_hadr_availability_replica_states  where synchronization_health<>2)
begin
    print 'alwayson集羣同步異常，請查看'
end
---------------監控數據庫同步狀態
if exists(select * from sys.dm_hadr_database_replica_states  where  synchronization_state_desc NOT IN('SYNCHRONIZED','SYNCHRONIZED'))
begin
    print 'alwayson有數據庫同步異常，請查看'
end
if  exists(  select * from sys.dm_tcp_listener_states  where  state_desc not in('ONLINE'))
 begin
     print 'alwayson偵聽器異常，請查看'
 end
 -----------------主從延遲的監控

 IF EXISTS(
 SELECT ag.name AS ag_name, ar.replica_server_name AS ag_replica_server, dr_state.database_id as database_id,
dr_state.log_send_queue_size, is_ag_replica_local = CASE
WHEN ar_state.is_local = 1 THEN N'LOCAL'
ELSE 'REMOTE'
END ,
ag_replica_role = CASE
WHEN ar_state.role_desc IS NULL THEN N'DISCONNECTED'
ELSE ar_state.role_desc
END
FROM (( sys.availability_groups AS ag JOIN sys.availability_replicas AS ar ON ag.group_id = ar.group_id )
JOIN sys.dm_hadr_availability_replica_states AS ar_state ON ar.replica_id = ar_state.replica_id)
JOIN sys.dm_hadr_database_replica_states dr_state on
ag.group_id = dr_state.group_id and dr_state.replica_id = ar_state.replica_id where dr_state.log_send_queue_size>300
)
 begin
     print 'alwayson主體有超過300M的日誌沒有同步，請儘快查看'
 end

 if exists(
 SELECT ag.name AS ag_name, ar.replica_server_name AS ag_replica_server, dr_state.database_id as database_id,
dr_state.redo_queue_size, is_ag_replica_local = CASE
WHEN ar_state.is_local = 1 THEN N'LOCAL'
ELSE 'REMOTE'
END ,
ag_replica_role = CASE
WHEN ar_state.role_desc IS NULL THEN N'DISCONNECTED'
ELSE ar_state.role_desc
END
FROM (( sys.availability_groups AS ag JOIN sys.availability_replicas AS ar ON ag.group_id = ar.group_id )
JOIN sys.dm_hadr_availability_replica_states AS ar_state ON ar.replica_id = ar_state.replica_id)
JOIN sys.dm_hadr_database_replica_states dr_state on
ag.group_id = dr_state.group_id and dr_state.replica_id = ar_state.replica_id where dr_state.redo_queue_size>300
 )
  begin
     print 'alwayson副本有較多日誌未重作，請登陸查看'
 end
 ---------------主從狀態切換等監控
 DECLARE @MASTER VARCHAR(20)
--select @MASTER=a.replica_server_name,a.join_state,a.join_state_desc,b.is_local ,b.role,b.role_desc  from sys.dm_hadr_availability_replica_cluster_states a join sys.dm_hadr_availability_replica_states   b
--on a.replica_id=b.replica_id where b.role=1 --and a.replica_server_name='ALWAYSON241'
select @MASTER=a.replica_server_name  from sys.dm_hadr_availability_replica_cluster_states a join sys.dm_hadr_availability_replica_states   b
on a.replica_id=b.replica_id where b.role=1 --and a.replica_server_name='ALWAYSON241'
 print @MASTER
IF @MASTER IS NULL
BEGIN
    PRINT '主從同步異常'
END
ELSE IF @MASTER<>'ALWAYSON241'
BEGIN
    PRINT 'alwayson進行了切換，如今主庫是'+@master
END
------------------------監控alwayson數據庫狀態
if exists(select * from sys.dm_hadr_database_replica_states a where a.database_state<>0 or a.synchronization_state not in(1,2))
begin
    print  'alwayson有數據庫異常'
end


-----------發郵件預警
----declare @name nvarchar(222)
----declare @mailbody nvarchar(4000);
----declare @maillabel nvarchar(100);
----set @maillabel='做業失敗'
--EXEC msdb.dbo.sp_send_dbmail @profile_name   =   'sendmail', 
--                                                @recipients   =   '729975475@qq.com', 
--                                                @subject   =   'alwayson預警', 
--                                                @body   =   'alwayson故障

於alwayson相關的系統視圖主要涉及如下一些（這些系統的存儲過程能讓咱們很直觀的瞭解到咱們可用性組的狀態）：

--監控alwayson可用性管理器是否啓動
SELECT SERVERPROPERTY ( 'HadrManagerStatus' ) 
 --監控alwayson可用性組是否啓動
SELECT SERVERPROPERTY ( 'IsHadrEnabled' ) 
--查看服務器的集羣信息
SELECT * FROM sys.dm_hadr_cluster
SELECT * FROM sys.dm_hadr_cluster_members
select * from sys.dm_hadr_availability_replica_cluster_nodes
SELECT * FROM sys.dm_hadr_name_id_map

-----------------------------------------------
SELECT * FROM sys.availability_groups
SELECT * FROM sys.availability_groups_cluster
SELECT * FROM sys.dm_hadr_availability_group_states

--監控可用性副本
select * from sys.availability_replicas
--查看只讀路由
select * from sys.availability_read_only_routing_lists
---監控可用性副本狀態
select * from sys.dm_hadr_availability_replica_cluster_states
select * from sys.dm_hadr_availability_replica_states  --synchronization_health的狀態爲2纔算正常。須要監控
--監視可用性數據庫
select * from sys.availability_databases_cluster 
select * from sys.databases where name in(select database_name from  sys.availability_databases_cluster)
--該視圖包含對應於給定主/輔助數據庫上最新自動頁修復嘗試的行，每一個數據庫最多可對應 100 行,若是存儲數據就要手動查看
select * from sys.dm_hadr_auto_page_repair
--除了已同步和爲同步的alwayson的狀態
select * from sys.dm_hadr_database_replica_states  where  synchronization_state_desc NOT IN('SYNCHRONIZED','SYNCHRONIZED')

 -------------------------監視可用性組偵聽器的監控
 select * from sys.availability_group_listener_ip_addresses  where state_desc='ONLINE'
 select * from sys.availability_group_listeners
 -----咱們能夠用 NOT EXISTS來判斷
 select dns_name,port,ip_address from  sys.availability_group_listener_ip_addresses A JOIN sys.availability_group_listeners B 
 ON a.listener_id=b.listener_id  where a.state_desc='ONLINE'
 
 ---------------------------------------
 select * from sys.dm_tcp_listener_states  where  state_desc not in('ONLINE')