fuel health ceilometer fails because it loads all resources

[Fabrizio Soppelsa]

6.0

How to reproduce:

- Deploy a HA environment with Telemetry roles, do some operations
- Run fuel health --env <ENVID> --check sanity

Result:
[ 1 of 15] [failure] 'List ceilometer availability' (77.61 s) Time limit exceeded while waiting for Resource listing. to finish. Please refer to OpenStack logs for more details.

Expected:
[ 1 of 15] [success]

I haven't read the internals, but maybe we can restrict the selection of resources to make this health check instead than listing all the resources. Like querying:
ceilometer resource-list -q “timestamp>2015-01-01T00:00:00” -q “resource_id=<resource id>”

[Vadim Rovachev (vrovachev)]

In 7.0 this implemented https://github.com/stackforge/fuel-ostf/blob/master/fuel_health/tests/sanity/test_sanity_ceilometer.py#L46

[Vitaly Sedelnik (vsedelnik)]

Won't Fix for 6.0-updates because this issue has no real impact

[Roman Rufanov (rrufanov)]

Customer found on 6.0. Reopening since need to include in MU for 6.0 and 6.1

[Ilya Tyaptin (ityaptin)]

Hi, Vitaly and Roman.

This bug affects only Fuel 6.0, in 6.1 and 7.0 ostf tests load only resources by one hour [1][2]. Please, mark it as invalid for 6.1, 7.0.

Vitaly, in 6.0 this bug have not real impact but ostf health checks for ceilometer fails on longliving environments even if ceilometer works. I think it will be cool to fix this wrong behavior.

[1] https://github.com/stackforge/fuel-ostf/blob/stable/6.1/fuel_health/tests/sanity/test_sanity_ceilometer.py#L48
[2] https://github.com/stackforge/fuel-ostf/blob/stable/7.0/fuel_health/tests/sanity/test_sanity_ceilometer.py#L48

[Vitaly Sedelnik (vsedelnik)]

Invalid for 7.0-updates and 6.1-updates per comment above. Won't Fix for 6.0-updates as there is no customer impact