Change the `RetryFunc` from a plain `error` return type to a
specialized `RetryError` which must decide whether it is
retryable or not.
Add `RetryableError` / `NonRetryableError` factory functions that
callers are meant to use to build up these errors.
This makes it eminently clear whether or not a given error is
retryable from inside the client code.
Goal here is to _not_ change any behavior, simply reflect the
existing behavior with the new, clearer, API.
Fixes a diff calculation error when only a VPC zone
identifiers is provided. In this case the associated
availability zones are computed from the subnets per
the AWS documentation.
It was a mistake to switched fully to `==` when activating waiting for
capacity on updates in #3947. Users that didn't set `min_elb_capacity ==
desired_capacity` and instead treated it as an actual "minimum" would
see timeouts for every create, since their target numbers would never be
reached exactly.
Here, we fix that regression by restoring the minimum waiting behavior
during creates.
In order to preserve all the stated behavior, I had to split out
different criteria for create and update, criteria which are now
exhaustively unit tested.
The set of fields that affect capacity waiting behavior has become a bit
of a mess. Next major release I'd like to rework all of these into a
more consistently named block of config. For now, just getting the
behavior correct and documented.
(Also removes all the fixed names from the ASG tests as I was hitting
collision issues running them over here.)
Fixes#4792
It's a bit confusing to have Terraform poll until instances come up on
ASG creation but not on update. This changes update to also poll if
min_size or desired_capacity are changed.
This changes the waiting behavior to wait for precisely the desired
number of instances instead of that number as a "minimum". I believe
this shouldn't have any undue side effects, and the behavior can still
be opted out of by setting `wait_for_capacity_timeout` to 0.
The `ForceDelete` parameter was getting sent to the upstream API call,
but only after we had already finished draining instances from
Terraform, so it was a moot point by then.
This fixes that by skipping the drain when force_delete is true, and it
also simplifies the field config a bit:
* set a default of false to simplify the logic
* remove `ForceNew` since there's no need to replace the resource to
flip this value
* pull a detail comment from code into the docs
The initial commit of AWS autoscaling group termination policy was
unfinished. It only worked on "create", and so had a needless ForceNew
that would rebuild autoscaling groups on any change. It also used a
HashString set, so it didn't preserve ordering of multiple policies
correctly.
Added the "update" operation, and converted to a TypeList to preserve
ordering. In addition, removing the policy or setting it to a null list
will reset the policy to "Default", the standard AWS policy.
Updated the acceptance tests to verify the update, but the null case is
difficult to test.
This landed in aws-sdk-go yesterday, breaking the AWS provider in many places:
3c259c9586
Here, with much sedding, grepping, and manual massaging, we attempt to
catch Terraform up to the new `awserr.Error` interface world.
If an AutoScalingGroup is in the middle of performing a Scaling
Activity, it cannot be deleted, and yields a ScalingActivityInProgress
error.
Retry the delete for up to 5m so we don't choke on this error. It's
telling us something's in progress, so we'll keep trying until the
scaling activity completed.
On ASG creation, waits for up to 10m for desired_capacity or min_size
healthy nodes to show up in the group before continuing.
With CBD and proper HealthCheck tuning, this allows us guarantee safe
ASG replacement.
Mitchell Hashimoto