title: SQLServer · CASE分析 · 镜像状态异常排查

author: 天铭

问题

用户实例的某个DB一直处于Synchronizing无法达到SYNCHRONIZED状态，用了很多修复方法并且进行了镜像重搭，但依旧没有达到正常同步态

排查

Synchronizing态通常和redo queue或者send queue有关

主库的问题DB当前的TPS在3K左右

begin tran
DECLARE @value int
DECLARE @value2 int
select @value=CONVERT(int,cntr_value) from sys.dm_os_performance_counters
where object_name like '****:database%%'and instance_name='businesscard' and counter_name like 'Transactions/    sec%%'
waitfor delay '00:00:01'
select @value2=CONVERT(int,cntr_value) from sys.dm_os_performance_counters
where object_name like '****:database%%'and instance_name='businesscard' and counter_name like 'Transactions/    sec%%'
select @value2-@value
commit tran

主库的发送队列

begin tran
DECLARE @value int
DECLARE @value2 int
select @value=CONVERT(int,cntr_value) from sys.dm_os_performance_counters
where counter_name = 'Log Send Queue KB' and instance_name='businesscard';
waitfor delay '00:00:01'
select @value2=CONVERT(int,cntr_value) from sys.dm_os_performance_counters
where counter_name = 'Log Send Queue KB' and instance_name='businesscard';
select @value*1./1024 as first_second_MB,@value2*1./1024 as second_second_MB,(@value2-@value)*1./1024 as diff_MB
commit tran

备库的应用队列

begin tran
DECLARE @value int
DECLARE @value2 int
select @value=CONVERT(int,cntr_value) from sys.dm_os_performance_counters
where counter_name = 'Redo Queue KB' and instance_name='businesscard'
waitfor delay '00:00:01'
select @value2=CONVERT(int,cntr_value) from sys.dm_os_performance_counters
where counter_name = 'Redo Queue KB' and instance_name='businesscard';
select @value*1./1024 as first_second_MB,@value2*1./1024 as second_second_MB,(@value2-@value)*1./1024 as diff_MB
commit tran

备库的应用速度大概在 14MB

begin tran
DECLARE @value bigint
DECLARE @value2 bigint
select @value=CONVERT(bigint,cntr_value)*1./1024/1024 from sys.dm_os_performance_counters
where counter_name = 'Redo Bytes/sec' and instance_name='businesscard';
waitfor delay '00:00:01'
select @value2=CONVERT(bigint,cntr_value)*1./1024/1024 from sys.dm_os_performance_counters
where counter_name = 'Redo Bytes/sec' and instance_name='businesscard';
select (@value2-@value) as speed_MB
commit tran

假设忽略主库新产生的日志，追上主库需要的时间 32min
select 27338.278320/14/60 =32.545569428566
如果再算上主库新产生的大概在40min左右

查看其它counter值

select cntr_value,* from sys.dm_os_performance_counters
where counter_name in(
'Log Send Flow Control Time (ms)','Bytes Sent/sec','Log Bytes Sent/sec',
'Log Compressed Bytes Sent/sec','Log Harden Time (ms)','Log Send Flow Control Time (ms)',
'Log Send Queue KB','Mirrored Write Transactions/sec','Pages Sent/sec',
'Send/Receive Ack Time','Sends/sec','Transaction Delay' )
and instance_name='businesscard';

Send/Receive Ack Time：
Milliseconds that messages waited for acknowledgement from the partner, in the last second.
This counter is helpful in troubleshooting a problem that might be caused by a network bottleneck, such as unexplained failovers, a large send queue, or high transaction latency. In such cases, you can analyze the value of this counter to determine whether the network is causing the problem.

开始怀疑问题在网络上，该主机某个网络组建已经升级了，但是备机的组件未升级
等待确认的时间稳定在800ms左右，对比主备网络组件都升级的主机在100ms左右，这个值跟事务大小有关，目前发现有问题的地方可能是这里

begin tran
DECLARE @value bigint
DECLARE @value2 bigint
select @value=CONVERT(bigint,cntr_value) from sys.dm_os_performance_counters
where counter_name = 'Send/Receive Ack Time' and instance_name='_Total';
waitfor delay '00:00:01'
select @value2=CONVERT(bigint,cntr_value) from sys.dm_os_performance_counters
where counter_name = 'Send/Receive Ack Time' and instance_name='_Total';
select (@value2-@value) as 'Send/Receive Ack Time'
commit tran

另外查看errorlog，时有告警镜像的状态应该是介于suspended和SYNCHRONIZING之间，错误类似KB,2008r2 应该已经修复。

半小时后发送队列又变大了

备库的应用队列减小的也非常慢

考虑单点时间过久，只能重搭解决，但重搭后追日志还是出现此问题，怀疑可能有坏页或者其它未知情况，计划在维护时间做checkdb。

--- ---
checkdb发现没有坏页，说明数据是ok的，那么问题可能在日志了。

select log_reuse_wait,log_reuse_wait_desc,* from sys.databases where name='businesscard'

log_reuse_wait_desc 为 ACTIVE_TRANSACTION

ACTIVE_TRANSACTION：事务处于活动状态。
一个长时间运行的事务可能存在于日志备份的开头。在这种情况下，可能需要进行另一个日志备份才能释放空间。
事务被延迟（仅适用于 SQL Server 2005 Enterprise Edition 及更高版本）。“延迟的事务”是有效的活动事务，因为某些资源不可用，其回滚受阻。

dbcc opentran('businesscard')

15号的事务今天19号，导致中间的日志全是活动日志无法截断，跟用户确认kill掉后，再次查看活动日志

dbcc opentran('businesscard')

发现结果一直在变化，可以理解为活动日志一直在往前走（越来越少），再次备份后活动日志恢复到7G左右，重搭成功。

事实上这是很早以前处理的一个CASE，当时排查了很久才找到root cause但现在看来问题其实很简单，希望大家下次遇到可以很快处理好