RFC: #786844 - Detection of deleted instances

Mo Morsi mmorsi at redhat.com
Thu Feb 9 15:37:51 UTC 2012


On 02/09/2012 10:05 AM, Matt Wagner wrote:
> On Thu, Feb 09, 2012 at 09:53:53AM -0500, Mo Morsi wrote:
>> How configurable can we make this delay? Eg can we allow the admin to
>> change the granularity of the poll interval? Or the number of failed
>> attempts before deleting the record? Or is this done on a per-request
>> basis or such?
> We *could*, but I'm not sure I see the benefit. This whole approach is
> somewhat of a stop-gap until we can implement cleaner handling here, so
> giving people knobs to tweak these parameters now would just distract us
> from fixing the real underlying issue. I'll implement them as constants
> so we can at least tweak them at the top of the file if you really need
> to, as opposed to having magic numbers sprinkled throughout the code.
>
> -- Matt

Hrm was more curious as to when conductor actually queries deltacloud to 
update instance state and other attributes?

Again this ultimately boils down to keeping the overall problem of 
keeping the cloud data that is cached in the local db in sync w/ the 
actual data provided by the cloud.

   -Mo



More information about the aeolus-devel mailing list