Bump timeout for grafana startup

The initial migrations when starting grafana for the first time may
sometimes take much longer than 20s, we have seen samples up to near
60s. Allow 120s to have some margin. Also make the timeout parameters
configurable.

Closes-Bug: 1769962
Signed-off-by: Dr. Jens Harbott <[email protected]>
Change-Id: If9186d8aa65150c492657550064789e211dbb570
(cherry picked from commit f8f34e0)

Author: osfrickler

ansible/roles/grafana/defaults/main.yml

@@ -81,6 +81,9 @@ grafana_default_volumes:
   - "kolla_logs:/var/log/kolla/"
 grafana_extra_volumes: "{{ default_extra_volumes }}"
 
+grafana_start_first_node_delay: 10
+grafana_start_first_node_retries: 12
+
 ############
 # Prometheus
 ############

ansible/roles/grafana/handlers/main.yml

@@ -29,8 +29,8 @@
       status_code: 200
   register: result
   until: result.get('status') == 200
-  retries: 10
-  delay: 2
+  retries: "{{ grafana_start_first_node_retries }}"
+  delay: "{{ grafana_start_first_node_delay }}"
   when:
     - kolla_action != "config"
     - inventory_hostname == groups[service.group]|first

releasenotes/notes/grafana-start-first-node-timeout-f9a6149cc68153a5.yaml

@@ -0,0 +1,10 @@
+---
+fixes:
+  - |
+    On slower nodes, the initial grafana startup could experience a
+    timeout failure when the migrations for setting up the database
+    took longer than expected. This has been fixed by increasing the
+    default timeout. The timeout settings can be changed via new
+    parameters ``grafana_start_first_node_delay`` and
+    ``grafana_start_first_node_retries`` for the ``grafana`` role.
+    `LP#1769962 <https://launchpad.net/bugs/1769962>`__