用于Windows 2016集群磁盘设备的LIO iSCSI目标尚未准备好

Question

我在一台Fedora 22机器上配置了一个LIO iSCSI目标，它将1 LUN导出到2台Windows 2016机器，这两台机器都能够无问题地访问磁盘。

这些服务器是HA集群的一部分，当我试图将磁盘添加到集群时，我会得到一个错误，说明“设备还没有准备好”。验证工具说，存储不会因为SCSI-3的持续保留而通过，但是从我的研究来看，这应该会给出一个与我所收到的不同的错误。

   Creating the physical disk resource for 'Cluster Disk 1'.

Bringing the resource for 'Cluster Disk 1' online.

There was an error creating, configuring, or bringing online the Physical Disk resource (disk) 'Cluster Disk 1'.

The following errors occurred while adding storage to the cluster:

The resource 'Cluster Disk 1' did not come online.

The desired state change for 'Cluster Disk 1' did not occur before the timeout expired.

这是我的iSCSI目标：

o- / ........................................................................................................... [...]
  o- backstores ................................................................................................ [...]
  | o- block .................................................................................... [Storage Objects: 1]
  | | o- ha1 ................................................ [/dev/delta/volpool/ha1 (200.0GiB) write-thru activated]
  | o- fileio ................................................................................... [Storage Objects: 0]
  | o- pscsi .................................................................................... [Storage Objects: 0]
  | o- ramdisk .................................................................................. [Storage Objects: 0]
  | o- user ..................................................................................... [Storage Objects: 0]
  o- iscsi .............................................................................................. [Targets: 1]
  | o- iqn.2017-12.net.hirstgroup.adx.delta:storage.target00 ............................................... [TPGs: 1]
  |   o- tpg1 ................................................................................. [no-gen-acls, no-auth]
  |     o- acls ............................................................................................ [ACLs: 4]
  |     | o- iqn.1991-05.com.microsoft:dc1.adx.hirstgroup.net ....................................... [Mapped LUNs: 1]
  |     | | o- mapped_lun0 ..................................................................... [lun0 block/ha1 (rw)]
  |     | o- iqn.1991-05.com.microsoft:dc2.adx.hirstgroup.net ....................................... [Mapped LUNs: 1]
  |     | | o- mapped_lun0 ..................................................................... [lun0 block/ha1 (rw)]
  |     | o- iqn.2017-12.net.hirstgroup.adx.dc1:dc1 ................................................. [Mapped LUNs: 1]
  |     | | o- mapped_lun0 ..................................................................... [lun0 block/ha1 (rw)]
  |     | o- iqn.2017-12.net.hirstgroup.adx.delta:iqn.1991-05.com.microsoft:dc2.adx.hirstgroup.net .. [Mapped LUNs: 1]
  |     |   o- mapped_lun0 ..................................................................... [lun0 block/ha1 (rw)]
  |     o- luns ............................................................................................ [LUNs: 1]
  |     | o- lun0 ............................................................... [block/ha1 (/dev/delta/volpool/ha1)]
  |     o- portals ...................................................................................... [Portals: 1]
  |       o- 0.0.0.0:3260 ....................................................................................... [OK]
  o- loopback ........................................................................................... [Targets: 0]
  o- vhost .............................................................................................. [Targets: 0]

所以，我不知道我做错了什么，似乎所有的事情都正常，除了磁盘不会添加到集群。在我的研究中，我确实看到了一件事，那就是SCSI-3的持续保留是起作用的必要条件，但根据我的理解，LIO确实支持这一点。当我在这台机器上运行ZFS时，我的块设备实际上是一个瘦配置的zvol。

希望有人能帮助澄清这里出了什么问题。

Answer 1

您可能需要尝试清除SCSI-3持久性预订，因为LIO可能会在LUN所有权更改后“忘记”更新它们。

清空-群集磁盘预订

Answer 2

我解决了这个问题。看来，Fedora 22中包含的LIO版本并不正确地支持SCSI-3的持久性保留。我转而使用scsi-target-utils，这个配置没有出现问题：

     backing-store /blah/blah/blah/zvol         initiator-address 172.16.20.0/24         incominguser hgx blahblahblah