snapshot an instance and delete the instance at same time lead to race condition

[nafeiyang]

Reported by CI, zCC will not delete the source image from DB after export, it will cause the zCC use out of space quickly if run snapshot related case.

2018-05-07 23:45:53.703 8160 ERROR nova.virt.zvm.utils [req-009743-451b-4095-bcfa-85ad4ba8ce84 tempest-ServerPasswordTestJSON-541812697 tempest-ServerPasswordTestJSON-541812697] zVM Cloud Connector request guest_capture failed with parameters: ('RMS0001a', u'2d523db1-0fe5-49af-aebd-20578eb8c2fc') {} . Results: {u'rs': 5, u'overallRC': 300, u'modID': 10, u'rc': 300, u'output': u'', u'errmsg': u'Failed to capture userid RMS0001a to generate image, error: \\nERROR: An error was encountered while creating disk image at stage(rc): ckdencode(5), gzip(1) gzip: stdout: No space left on device Error: ckdencode encountered an unexpected error writing to STDOUT, errno: 32'}

[nafeiyang (nafeiy)]

Fix committed
https://review.gerrithub.io/#/c/mfcloud/python-zvm-sdk/+/410342/

[jichenjc (jichenjc)]

actually this is because delete an instance when instance is doing snapshot
this leads to some issue because image might be captured in zcc but failed to be cleanup by driver

[jichenjc (jichenjc)]

https://bugs.launchpad.net/nova/+bug/1754071