(新手)关于虚拟机备份提示11、13的问题

A21270

本人新手。
目前进行NBU虚拟机备份的时候，一些客户端时常发现11错误（系统调用失败）或者是13错误（文件读取失败）

特征是：备份到一部分的时候，备份失败。例如400G的虚拟机备份到200G的时候就显示这个信息。

一般来说，会是空间不足的问题吗？还是虚拟空间的问题？具体有什么推荐调整的方式。

报case，显示以下信息：

1.vxms日志显示我们每次备份到那个数据量大小就中断的原因是：该主机每次都只能备份yzbbxt-db.vmdk 磁盘，而磁盘yzbbxt-db_1.vmdk在尝试读取3次失败后就退出了，即yzbbxt-db_1.vmdk磁盘没有正常读取。
日志如下：
Begin:vfm_read [19:02:39.0393 : ]
In: nreq=1
  vfm_obj_t {  orig_name = [HDS_HUS_LUN1_2TB] yzbbxt-db/yzbbxt-db.vmdk  }

  [ I/O Req ]
   vfm_io_req_t { offset = 214748270080 reqlen = 65536 bytes_trans = 0 bytes_req = 0 error = 0  }

Out:

  [ I/O Req  ]
   vfm_io_req_t { offset = 214748270080 reqlen = 65536 bytes_trans = 65536 bytes_req = 0 error = 0  } --214748270080是读取的数据量，会增长的

End0)

Begin:vfm_read [19:02:39.0395 : ]
In: nreq=1
  vfm_obj_t {  orig_name = [HDS_HUS_LUN1_2TB] yzbbxt-db/yzbbxt-db.vmdk  }

  [ I/O Req ]
   vfm_io_req_t { offset = 214748335616 reqlen = 29184 bytes_trans = 0 bytes_req = 0 error = 0  }

Out:

  [ I/O Req  ]
   vfm_io_req_t { offset = 214748335616 reqlen = 29184 bytes_trans = 29184 bytes_req = 0 error = 0  }

End0)

Begin:vfm_close [19:02:39.0397 : ]    ---关闭读取
In:  vfm_obj_t {  orig_name = [HDS_HUS_LUN1_2TB] yzbbxt-db/yzbbxt-db.vmdk  }   ---这是这个盘读取完成的标记

End0)   --正常返回

Begin:vfm_open_file_name [19:02:39.0431 : ]
In: filename=<vix>[HDS_HUS_LUN1_2TB] yzbbxt-db/yzbbxt-db_1.vmdk   ---开始读取下一块盘
flags=2
context=

Out:
rflags=0
  vfm_obj_t {  orig_name = [HDS_HUS_LUN1_2TB] yzbbxt-db/yzbbxt-db_1.vmdk  }

End0)

Begin:vfm_read [19:02:39.0432 : ]
In: nreq=1
  vfm_obj_t {  orig_name = [HDS_HUS_LUN1_2TB] yzbbxt-db/yzbbxt-db_1.vmdk  }

  [ I/O Req ]
   vfm_io_req_t { offset = 0 reqlen = 35328 bytes_trans = 0 bytes_req = 0 error = 0  }   ---没有读取到数据，返回11错误，这种操作尝试了3次都失败。

End11)

2.bpbkar日志中有明确的读取3次失败的记录
19:03:02.806: [1604.568] <2> tar_base::V_vTarMsgW: TRV - Read of VM file read 0 bytes, should have read 35328 bytes, will be issuing retry 1 of 3 (../flashmap.cpp:2006)
19:03:03.806: [1604.568] <2> tar_base::V_vTarMsgW: TRV - VxMS Error message 1 = fiml_read_file vfm_read
message 2 = stat 11 req.io_error 0 (../vxms_bridge.cpp:320)   ----bkbkar日志在读取文件时报11错误。
19:03:03.806: [1604.568] <2> tar_base::V_vTarMsgW: TRV - Read of VM file read 0 bytes, should have read 35328 bytes, will be issuing retry 2 of 3 (../flashmap.cpp:2006)
.....
19:03:04.807: [1604.568] <2> tar_base::V_vTarMsgW: TRV - Read of VM file read 0 bytes, should have read 35328 bytes, will be issuing retry 3 of 3 (../flashmap.cpp:2006)

以下为ESX server的日志：

1.您这ESX server的时间和北京时间有时差，不过时差是一天内不算问题
2.vpxa的日志中针对disk的update，我给你做了如下2端日志的对比，明显显示，opid如果能够正常更新就会正常退出，但是opid SWI-9e91f12e一直更新不过去，请检查。

正常日志，会很快update：
2016-09-14T06:37:07.128Z [550C7B70 verbose 'VpxaHalCnxHostagent' opID=WFU-91f9690] [WaitForUpdatesDone] Received callback
2016-09-14T06:37:07.128Z [550C7B70 verbose 'VpxaHalCnxHostagent' opID=WFU-91f9690] [VpxaHalCnxHostagent:rocessUpdate] Applying updates from 5434414 to 5434415 (at 5434414)
2016-09-14T06:37:07.128Z [550C7B70 verbose 'hostdvm' opID=WFU-91f9690] [VpxaHalVmHostagent] 34: GuestInfo changed 'guest.disk'
2016-09-14T06:37:07.129Z [550C7B70 verbose 'halservices' opID=WFU-91f9690] [VpxaHalServices] VmGuestDiskChange Event for vm(40) 34
2016-09-14T06:37:07.129Z [550C7B70 verbose 'vpxavpxaInvtVm' opID=WFU-91f9690] [VpxaInvtVmChangeListener] Guest DiskInfo Changed
2016-09-14T06:37:07.129Z [550C7B70 verbose 'VpxaHalCnxHostagent' opID=WFU-91f9690] [WaitForUpdatesDone] Starting next WaitForUpdates() call to hostd
2016-09-14T06:37:07.129Z [550C7B70 verbose 'VpxaHalCnxHostagent' opID=WFU-91f9690] [WaitForUpdatesDone] Completed callback

opID=SWI-9e91f12e  更新不成功。（这个opid你得找vmware去做对应，我只知道这个opid  update失败）。
2016-09-14T06:46:26.181Z [FFB03B70 verbose 'vpxavpxaMoVm' opID=SWI-9e91f12e] [VpxaMoVm::CheckMoVm] did not find a VM with ID 59 in the vmList
2016-09-14T06:46:26.181Z [FFB03B70 verbose 'vpxavpxaAlarm' opID=SWI-9e91f12e] [VpxaAlarm] VM with vmid = 59 not found
2016-09-14T06:46:36.183Z [FFB03B70 verbose 'vpxavpxaMoVm' opID=SWI-9e91f12e] [VpxaMoVm::CheckMoVm] did not find a VM with ID 59 in the vmList
2016-09-14T06:46:36.183Z [FFB03B70 verbose 'vpxavpxaAlarm' opID=SWI-9e91f12e] [VpxaAlarm] VM with vmid = 59 not found

3.hostd.log日志中关于这个磁盘yzbbxt-db_1.vmdk有error ，请找vmware去分析。
2016-09-14T03:34:27.003Z [2A681B70 verbose 'Default' opID=69c6fd79-a9 user=vpxuser] AdapterServer: target='vim.VirtualMachine:37', method='queryChangedDiskAreas'
2016-09-14T03:34:27.017Z [2A681B70 error 'Vmsvc.vm:/vmfs/volumes/5036aacb-761921f0-4ceb-3440b5b0b762/yzbbxt-db/yzbbxt-db.vmx' opID=69c6fd79-a9 user=vpxuser] Cannot compute changes for disk /vmfs/volumes/5036aacb-761921f0-4ceb-3440b5b0b762/yzbbxt-db/yzbbxt-db_1.vmdk: Unknown change epoch.
2016-09-14T03:34:27.017Z [2A681B70 info 'Default' opID=69c6fd79-a9 user=vpxuser] AdapterServer caught exception: vim.fault.FileFault
2016-09-14T03:34:27.017Z [2A681B70 info 'Solo.Vmomi' opID=69c6fd79-a9 user=vpxuser] Activation [N5Vmomi10ActivationE:0x28694ca0] : Invoke done [queryChangedDiskAreas] on [vim.VirtualMachine:37]

4.该ESX server 是否使用了Vcenter，建议把backup hosts主机的IP 地址解析也添加到 ESX server上。