sql-server - SQL 2017 Linux ag 资源未通过pacemaker 进行故障转移

Question

我们已经在 Linux 上设置了一个 sql2017 集群，如下 MS Documentation 。 AG 中的复制工作正常，但我们无法进行故障转移。如果我在故障转移期间查看日志，pacemaker 正在尝试移动 AG，但它失败并继续在主服务器上运行。

在主服务器上，它报告资源未运行。

Oct 01 15:06:56 [4346] syncdb01a-stag       lrmd:   notice: operation_finished: ttsyncagresource_monitor_11000:6280:stderr [ resource ttsyncagresource is NOT running ]

在辅助设备上我看到这个未知错误:

Oct 01 15:06:57 [24662] syncdb01b-stag  pengine:  warning: unpack_rsc_op_failure:   Processing failed start of ttsyncagresource:1 on syncdb01b-stag: unknown error | rc=1

如果我运行pcs status，我会得到以下信息。它显示的最新错误是如果我关闭主节点会发生什么。另外两个错误是由于 sql 权限造成的，已解决。

[root@syncdb01a-stag oper]# pcs status
Cluster name: syncdb-stag
Stack: corosync
Current DC: syncdb01b-stag (version 1.1.20-5.el7_7.1-3c4c782f70) - partition with quorum
Last updated: Tue Oct  1 20:36:32 2019
Last change: Tue Oct  1 15:53:57 2019 by root via crm_resource on syncdb01a-stag

3 nodes configured
3 resources configured

Online: [ syncdb01a-stag syncdb01b-stag syncwit01-stag ]

Full list of resources:

 Master/Slave Set: ttsyncagresource-master [ttsyncagresource]
     Masters: [ syncdb01a-stag ]
     Stopped: [ syncdb01b-stag syncwit01-stag ]

Failed Resource Actions:
* ttsyncagresource_monitor_11000 on syncdb01a-stag 'not running' (7): call=17, status=complete, exitreason='',
    last-rc-change='Tue Oct  1 15:03:47 2019', queued=0ms, exec=0ms
* ttsyncagresource_start_0 on syncdb01b-stag 'unknown error' (1): call=17, status=complete, exitreason='2019/10/01 14:43:30 Did not find AG row in sys.availability_groups',
    last-rc-change='Tue Oct  1 14:43:25 2019', queued=0ms, exec=5255ms
* ttsyncagresource_start_0 on syncwit01-stag 'unknown error' (1): call=17, status=complete, exitreason='2019/10/01 14:43:30 Did not find AG row in sys.availability_groups',
    last-rc-change='Tue Oct  1 14:43:25 2019', queued=1ms, exec=5228ms

Daemon Status:
  corosync: active/enabled
  pacemaker: active/enabled
  pcsd: active/enabled

我还删除了任何限制(由于是多子网，我们没有使用虚拟 IP)

[root@syncdb01a-stag oper]# pcs constraint
Location Constraints:
Ordering Constraints:
Colocation Constraints:
Ticket Constraints:

这是pcs config的输出:

[root@syncdb01a-stag oper]# pcs config
Cluster Name: syncdb-stag
Corosync Nodes:
 syncdb01a-stag syncdb01b-stag syncwit01-stag
Pacemaker Nodes:
 syncdb01a-stag syncdb01b-stag syncwit01-stag

Resources:
 Master: ttsyncagresource-master
  Meta Attrs: notify=true
  Resource: ttsyncagresource (class=ocf provider=mssql type=ag)
   Attributes: ag_name=ttsyncag
   Meta Attrs: failure=timeout=60s notify=true
   Operations: demote interval=0s timeout=10 (ttsyncagresource-demote-interval-0s)
               monitor interval=10 timeout=60 (ttsyncagresource-monitor-interval-10)
               monitor interval=11 role=Master timeout=60 (ttsyncagresource-monitor-interval-11)
               monitor interval=12 role=Slave timeout=60 (ttsyncagresource-monitor-interval-12)
               notify interval=0s timeout=60 (ttsyncagresource-notify-interval-0s)
               promote interval=0s timeout=60 (ttsyncagresource-promote-interval-0s)
               start interval=0s timeout=60 (ttsyncagresource-start-interval-0s)
               stop interval=0s timeout=10 (ttsyncagresource-stop-interval-0s)

Stonith Devices:
Fencing Levels:

Location Constraints:
Ordering Constraints:
Colocation Constraints:
Ticket Constraints:

Alerts:
 No alerts defined

Resources Defaults:
 No defaults set
Operations Defaults:
 No defaults set

Cluster Properties:
 cluster-infrastructure: corosync
 cluster-name: syncdb-stag
 cluster-recheck-interval: 2min
 dc-version: 1.1.20-5.el7_7.1-3c4c782f70
 have-watchdog: false
 start-failure-is-fatal: true
 stonith-enabled: false

Quorum:
  Options:

Answer 1

我从头开始重建了集群，它工作得很好，不太确定我哪里出了问题，但这次我在开始之前进行了完整的系统更新。