openstack/kolla.git/docker/nova/nova-api/extend_start.sh, branch stein-eol

create missing apache2 directory on Debian/Ubuntu

2019-12-31T02:03:08+00:00

Ubuntu/source deployment of several images (horizon, placement-api, zun) failed with: + exec /usr/sbin/apache2 -DFOREGROUND apache2: Syntax error on line 80 of /etc/apache2/apache2.conf: DefaultRuntimeDir must be a valid directory, absolute or relative to ServerRoot Change-Id: Ie2a1077f7def0743f1403341985e2109aa490026 (cherry picked from commit 932f09bcd65e15524b5229fdb6bc73a7e0111b27)

Run nova api_db sync before db sync

2019-06-27T14:24:57+00:00

Backport: Stein, Rocky Since cells v2 was introduced, nova operators must run two commands to migrate the database schemas of nova's databases - nova-manage api_db sync and nova-manage db sync. It is necessary to run them in this order, since the db sync may depend on schema changes made to the api database in the api_db sync. Executing the db sync first may fail, for example with the following seen in a Queens to Rocky upgrade: nova-manage db sync ERROR: Could not access cell0. Has the nova_api database been created? Has the nova_cell0 database been created? Has "nova-manage api_db sync" been run? Has "nova-manage cell_v2 map_cell0" been run? Is [api_database]/connection set in nova.conf? Is the cell0 database connection URL correct? Error: (pymysql.err.InternalError) (1054, u"Unknown column 'cell_mappings.disabled' in 'field list'") [SQL: u'SELECT cell_mappings.created_at AS cell_mappings_created_at, cell_mappings.updated_at AS cell_mappings_updated_at, cell_mappings.id AS cell_mappings_id, cell_mappings.uuid AS cell_mappings_uuid, cell_mappings.name AS cell_mappings_name, cell_mappings.transport_url AS cell_mappings_transport_url, cell_mappings.database_connection AS cell_mappings_database_connection, cell_mappings.disabled AS cell_mappings_disabled \nFROM cell_mappings \nWHERE cell_mappings.uuid = %(uuid_1)s \n LIMIT %(param_1)s'] [parameters: {u'uuid_1': '00000000-0000-0000-0000-000000000000', u'param_1': 1}] (Background on this error at: http://sqlalche.me/e/2j85) Despite this error, the command actually exits zero, so kolla does not notice the failure. This change reorders these two commands to run the api_db sync before the db sync. Change-Id: I0514c02aaee1a413a7979c3a426e0484df839861 Closes-Bug: #1832860 (cherry picked from commit 84a07102e2db92333ce416a520151f1a1856f162)

Fix nova-api extend_start.sh

2018-06-11T06:59:52+00:00

Patch https://review.openstack.org/#/c/558765 added an extra closure bracket }, this patch fixes it. Change-Id: I379ff17856509c9321b86c13a72eacc18f5c1202 Closes-Bug: #1775698

Apply Nova rolling upgrade logic

2018-06-05T09:54:37+00:00

This patchset apply Nova rolling upgrade logic [1] [1] https://docs.openstack.org/nova/latest/user/upgrade.html Change-Id: I2cc973896d08988c0db121673d357fd1d37e6e6c Implements: blueprint apply-service-upgrade-procedure

Fix if $user is root at extend_start

2017-05-12T12:40:52+00:00

In ubuntu is not evaluating correctly if $user == root check at extend_start. Changing to "$(whoami)" == 'root' fixes the issue Change-Id: I296a1f98764cdfa09650c483bc7beba53e6bdba4 Closes-Bug: #1690360

Add missing cleanup step for new apache images

2017-04-20T08:17:38+00:00

A few new images got apache as part of blueprint apache-packages-for-apis without a cleaning up. Without this step the containers may fail to restart due to runtime files already present. This commit fixes it for all images with the exception of heat-base so that it can be cleanly backported to stable/ocata. It also removes the unused APACHE_DIR variable from freezer-api and keystone extend_start.sh script. Change-Id: I47d75221b3d1aae91f11d3e0da83e221a00315b7 Closes-Bug: #1679565

Adds migration step from nova to nova_api database

2016-04-05T07:50:01+00:00

This patch adds a missing required step to migrate information from the nova to the new nova_api database. Closes-Bug: #1565497 Change-Id: Ia1554e12c688e5a6509af2d37f1825fd4ab666be Co-Authored-By: Steven Dake

Remove duplicate code in nova extend_start.sh

2016-02-23T07:58:46+00:00

Change-Id: I37d737509bc7cf3dc74bb13ecb0a6ae68080556e Partial-Bug: #1546944

Make Heka collect Nova logs

2016-02-19T21:49:18+00:00

Partially implements: blueprint heka Change-Id: I67dfc5d762d74454e4a852dd61c57fe9cd4b78c8

Create nova-api database

2016-02-02T21:56:54+00:00

This is a required patch. Kilo introduced a nova-api database but didn't use it. [1] Mitaka now uses it and has broken booting instances. [2] [1] https://wiki.openstack.org/wiki/ReleaseNotes/Kilo#OpenStack_Compute_.28Nova.29 [2] https://github.com/openstack/nova/blob/df0fca62cf5324f5b6eae0fed1f88c6c9ed61c71/releasenotes/notes/request-spec-api-db-b9cc6e0624d563c5.yaml Change-Id: If0a7f4ba4937fab8d60f5236ca00db06faa768c7 Closes-Bug: #1540579