Fuel 4.0 ostf platform_tests: "Time limit exceeded" even when "Currently in CREATE_FAILED status"

[Aleksandr Didenko]

fuel-4.0 134 iso
{"release": "4.0", "nailgun_sha": "66119c3f5010143121b4cb5ca4a0e9e2c69aeb52", "ostf_sha": "bb0802637abc40d2caa734756544033951739f09", "astute_sha": "7b77437a5d87dd5d06b45d152df80a681ffa25eb", "fuellib_sha": "80b5a396a8de5348f0a96ff15282a22df6a95a1a"}

Problem:
[platform_tests] Create stack, check its details, then update and delete Heat stack
Status: failure
Message: Time limit exceeded while waiting for stack status becoming 'CREATE_COMPLETE' to finish. Please refer to OpenStack logs for more details.

Platform health check works for 100 seconds and gives "Time limit exceeded" even when it gets CREATE_FAILED status.

Example from ostf-stdout.log:
fuel_health.heatmanager: DEBUG: Waiting for .... Currently in CREATE_FAILED status
fuel_health.test: DEBUG: Sleeping for 10 seconds

Snapshot attached.

[Mike Scherbakov (mihgen)]

Lowering priority to High. I want to avoid to many criticals - I believe if just one of OSTF tests fails because of the issue in test, then it should not be critical.
If whole OSTF framework doesn't work - it is critical. Or if deployment fails in one of the modes.

[Yuriy Yekovenko (yyekovenko)]

Fix for this issue was done as part of this commit: https://review.openstack.org/#/c/62323/

[Mike Scherbakov (mihgen)]

Please provide a separate patch with the fix.
Sorry, we can't merge your current patchset as we are behind of feature freeze date, and looks like you are adding functionality there.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/63093

[Yuriy Yekovenko (yyekovenko)]

commit sent to review.

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/63093
Committed: https://git.openstack.org/cgit/stackforge/fuel-ostf/commit/?id=b4a5ff53a7f24cb09eb00fbb553627c880ea5e0c
Submitter: Jenkins
Branch: master

commit b4a5ff53a7f24cb09eb00fbb553627c880ea5e0c
Author: Yuriy Yekovenko <email address hidden>
Date: Thu Dec 19 13:38:44 2013 +0200

    Fixed naming of expected status of a stack.

    Method wait_for_stack_status in heatmanager
    was waiting for "ERROR" status as reason
    for test interruption but should wait
    for "CREATE_FAILED" status.
    Similar fix was done in cleanup.py.

    Change-Id: Ib9fb5c3df3256328012d23a56bd4f78923724c6c
    Closes-Bug: #1260798