Implement Delete timeouts
* Implements new logic around when instances are deleted. See flowcharts in docs/. Basically timeouts now trigger to actually delete instances from a1s database, and these timeouts are configurable.
* Eliminates the barrier to deleting an instance when no xapp evdr replied (via timeouts)
* Add two new ENV variables that control timeouts
* Make unit tests more modular so new workflows can be tested easily
* Changes the API for ../status to return a richer structure
* Clean up unused items in the integration tests helm chart
* Removed "RMR_RCV_RETRY_INTERVAL" leftovers since this isn't used
anymore
Issue-ID: RICPLT-2619
Signed-off-by: Tommy Carpenter <tc677g@att.com>
Change-Id: I5b2f9cc3a6e8da7fe636a4cde085a5e1a3770f62
diff --git a/docs/installation-guide.rst b/docs/installation-guide.rst
index 5ddad48..babd4fb 100644
--- a/docs/installation-guide.rst
+++ b/docs/installation-guide.rst
@@ -1,5 +1,6 @@
.. This work is licensed under a Creative Commons Attribution 4.0 International License.
.. http://creativecommons.org/licenses/by/4.0
+.. Copyright (C) 2019 AT&T Intellectual Property
Installation Guide
==================
@@ -13,9 +14,11 @@
You can set the following ENVs to change the A1 behavior:
-1. ``RMR_RETRY_TIMES`` the number of times failed rmr operations such as
-timeouts and send failures should be retried before A1 gives up and
-returns a 503. The default is ``4``.
+1. ``A1_RMR_RETRY_TIMES``: the number of times failed rmr operations such as timeouts and send failures should be retried before A1 gives up and returns a 503. The default is ``4``.
+
+2. ``INSTANCE_DELETE_NO_RESP_TTL``: Please refer to the delete flowchart in docs/; this is ``T1`` there. The default is 5 (seconds). Basically, the number of seconds that a1 waits to remove an instance from the database after a delete is called in the case that no downstream apps responded.
+
+3. ``INSTANCE_DELETE_RESP_TTL``: Please refer to the delete flowchart in docs/; this is ``T2`` there. The default is 5 (seconds). Basically, the number of seconds that a1 waits to remove an instance from the database after a delete is called in the case that downstream apps responded.
K8S
---