heal_instance_info_cache_interval help text is inaccurate

[sean mooney]

the current help text for the config option is as follows

heal_instance_info_cache_interval¶

    Type

        integer
    Default

        60

    Interval between instance network information cache updates.

    Number of seconds after which each compute node runs the task of querying Neutron for all of its instances networking information, then updates the Nova db with that information. Nova will never update it’s cache if this option is set to 0. If we don’t update the cache, the metadata service and nova-api endpoints will be proxying incorrect network data about the instance. So, it is not recommended to set this option to 0.

    Possible values:

        Any positive integer in seconds.

        Any value <=0 will disable the sync. This is not recommended.

this is not correct.

when the value is set to 0 it will disable the periodic task so nova will not try to heal the cache but the cache will still get updated if neutron notifies us of a change.

every time neutron emits a network-changed event for an interface nova updates that interface in the info cache.

the reason that setting the value to 0 is not recommended has nothing to do with what the current help text states. it's not recommended because it is not tested today.

in large deployments with many compute agents it may be better for scaling to disable this periodic.

for environments that use ironic or a clustered hypervisor like vmware the default will not introduce a performance issue but it will take a very long time to heal all the nodes if there is an issue. that this would fix.

as a result it is likely better to diabel this longterm.

for now we should just fix the help text and add testing fo disabling this in a ci job.