Heat¶
This page is the starting point for Heat, the heat role: the containers heat_api, heat_api_cfn and heat_engine
(roles/heat/defaults/main.yml:4, roles/heat/defaults/main.yml:28, roles/heat/defaults/main.yml:52), on the
control nodes of the example inventory (inventory/multinode:128-129, inventory/multinode:350-357), selected by the
tag heat (site.yml:724-738). A changed heat_domain_admin_password reaches heat.conf but is never applied in
Keystone (roles/heat/templates/heat.conf.j2:12, kolla docker/heat/heat-api/extend_start.sh:10-14; still holds in
2024.1; Component dependencies §3). The lists below say where the study
covers how to operate it, its impact, its dependencies and the evidence, what to check around a run and what stays
open; the service card follows them.
What changed in 2024.1¶
- Container module renamed CODE: the three restart handlers, the check of the containers and the bootstrap
container now use
kolla_containerinstead ofkolla_docker, with the same actions (roles/heat/handlers/main.yml:2-48,roles/heat/tasks/check-containers.yml:2-5,roles/heat/tasks/bootstrap_service.yml:6-7). The handlers are the same three, in the same order, and still skipkolla_action == "config"(roles/heat/handlers/main.yml:15-16). - Public URLs and the external frontends CODE: the endpoints are built with the
kolla_urlfilter (roles/heat/defaults/main.yml:178-184). The public ones useheat_api_public_portandheat_api_cfn_public_port, which become 443 or 80 withhaproxy_single_external_frontend(defaultfalse) and otherwise stay 8004 and 8000 (group_vars/all.yml:412-419,group_vars/all.yml:931-932). The two external HAProxy entries now carryexternal_fqdnand the public port (roles/heat/defaults/main.yml:23-24,roles/heat/defaults/main.yml:47-48).heat.confbuildsheat_metadata_server_urlandheat_waitcondition_server_urlfrom the public CFN base endpoint (roles/heat/templates/heat.conf.j2:6-7), so turning the single external frontend on changesheat.confon every heat host, and the config task then notifies the restart of each heat container (roles/heat/tasks/config.yml:51-69). - New messaging options in
heat.confCODE: with the default quorum queues (group_vars/all.yml:268),rabbit_quorum_queue = trueis rendered (roles/heat/templates/heat.conf.j2:89-98; see F17). Withom_enable_queue_manager(defaultfalse,group_vars/all.yml:269) the three containers also mount/dev/shm(roles/heat/defaults/main.yml:154,roles/heat/defaults/main.yml:161,roles/heat/defaults/main.yml:168), andheat.confgetsuse_queue_manager, aprocessnameforheat-apiandheat-api-cfn(roles/heat/templates/heat.conf.j2:75-80) and an[oslo_concurrency]lock path (roles/heat/templates/heat.conf.j2:120-123). - Config validation CODE:
heat_config_validationnow points at/heat/etc/heat/heat-config-generator.conf(roles/heat/defaults/main.yml:64). - The play CODE: it still takes
serialfromkolla_serial, default0(site.yml:731), and now also setsmax_fail_percentage, default 100 (site.yml:732-735), which never stops the play early (Controlled batches §1.3). - Still holds in 2024.1 CODE: the image side of the bootstrap is unchanged: kolla
docker/heat/heat-api/extend_start.shis the same file in both tags (docker/heat/heat-api/extend_start.sh:5-17). The only kolla change for Heat is that the base image now copiesheat_api_audit_map.confinto/etc/heat/(kolladocker/heat/heat-base/Dockerfile.j2:30).
Where to look¶
Operate¶
No operation page gives a reconfigure command line for the heat tag. The scenario above describes
kolla-ansible -t <svc> reconfigure for an API service such as keystone, glance, cinder, heat or placement; its command
example is keystone (Reconfigure a single service
§1.1). The stop example of Stop and
destroy uses -t heat. For Heat, reconfigure runs deploy.yml
(roles/heat/tasks/reconfigure.yml:2): registration, config, check-containers, bootstrap, then the handlers
(roles/heat/tasks/deploy.yml:2-15) CODE.
Impact and risk¶
Dependencies¶
- Dependencies: Component dependencies §3
Evidence¶
- Fact sheet: fact sheet C §5
- Image side: Kolla images
§4 (row
heat-api), Kolla images §5
Before and after a run¶
- Tags §3.1, row
<svc>API service: safe on its own "yes, it is the intended unit; it still runs Keystone registration and DB sync"; batch withANSIBLE_SERIAL. Prerequisites: MariaDB, RabbitMQ, memcached, Keystone (registration), HAProxy VIP. - fact sheet C §5.8: with
serial 0, every heat-engine restarts in the same handler flush on any heat.conf change; a memcached or rabbit
membership change, a password rotation or an
openstack_service_workerschange is enough (site.yml:731;roles/heat/templates/heat.conf.j2:12,16,19,46,52,roles/heat/defaults/main.yml:196) CODE. Still holds in 2024.1. - R8, precautions for a HAProxy config change:
ANSIBLE_SERIAL=1for haproxy changes;--skip-tags haproxywhen you don't intend to change frontends. - R24, container stop timeout: in 2024.1 the stop goes through the container's systemd
unit, whose
ExecStopis/usr/bin/<engine> stop <name> -t <graceful_timeout>(module_utils/kolla_systemd_worker.py:32), withdocker_graceful_timeout60 s (group_vars/all.yml:139). Precautions: drain long operations before restarts; raisedocker_graceful_timeoutif needed (it reaches the unit on the next restart). The first 2024.1 check of a container whose unit 2023.1 wrote reports drift (F16). - Reconfigure a single service
§3,
verification of
-t <svc> reconfigure: before,kolla-ansible -t <svc> genconfig, thenkolla_set_configs --checkin each<svc>container and a diff of/etc/kolla/haproxy/services.d/<svc>*.cfgagainst the previous version; after,openstack endpoint list --service <name>is unchanged for each Keystone service the role registers, e.g.heatandheat-cfnfor heat, each with aninternaland apublicendpoint (roles/heat/defaults/main.yml:219-230), the API answers through the VIP, the container start times show which hosts restarted, and the haproxy container's start time shows whether HAProxy restarted. If the public port changed (see What changed in 2024.1), the public endpoints change too.
Known limits and open points¶
- Known bugs and limitations, row
heat:bootstrap_heatcreatesheat_user_domain/heat_domain_adminonly when the domain is absent: a changedheat_domain_admin_passwordreachesheat.confbut never Keystone. Status: CODE, still holds in 2024.1 (kolladocker/heat/heat-api/extend_start.sh:8-14,roles/heat/tasks/bootstrap_service.yml:22); impact NOT AVAILABLE. - Q14, answered from the source CODE: Heat creates
heat_user_domain/heat_domain_adminonly if the domain is absent, so a changedheat_domain_admin_passwordis never pushed to Keystone. - F15, "What the bootstrap containers run differs per image": Heat's bootstrap container
does not re-apply a changed
heat_domain_admin_passwordto Keystone CODE. - F3, "
docker_graceful_timeoutanddocker_client_timeoutare ignored": an Ussuri finding that no longer holds as written, in 2024.1 as in 2023.1. The role handlers passdocker_common_options, which carriesgraceful_timeout: "{{ docker_graceful_timeout }}"(60 s) andclient_timeout(120 s) (group_vars/all.yml:105,group_vars/all.yml:139,group_vars/all.yml:151-152), and the container's systemd unit stops it with<engine> stop <name> -t <graceful_timeout>(module_utils/kolla_systemd_worker.py:32). On every restart, a service still running 60 s after SIGTERM is killed with its in-flight work (heat-engine, cinder copies, glance uploads); the 10 s value is only the fallback when nograceful_timeoutis passed (library/kolla_container.py:370-371,module_utils/kolla_docker_worker.py:417-419,module_utils/kolla_docker_worker.py:441-443) CODE (R24). - F5, "
reconfigure=deploy": a reconfigure re-runs Keystone registration, DB creation/grants and the DB-sync bootstrap containers of most services CODE. - What is not
known, row
"Heat": the sources do not establish whether stack-domain operations fail authentication after a changed
heat_domain_admin_password. - What is not known: the sources do not establish whether xtrace writes one-shot secrets (the heat domain admin password among them) to the container's stderr.
- fact sheet C §5.8,
"Operational notes", four points: heat-engine restarts with stacks in progress (NOT AVAILABLE: upstream heat
behaviour); with serial 0, every heat-engine restarts in the same handler flush on any heat.conf change; on upgrade
the db sync runs while old engines are still active, and all engines restart together after it (2024.1, unchanged:
config, check-containers, then
bootstrap_service.yml, then the flush of handlers,roles/heat/tasks/upgrade.yml:1-9CODE); Heat does not touch compute hosts. - fact sheet C §8, items 8 and
10: NOT AVAILABLE: heat-engine restart with
CREATE_IN_PROGRESSstacks: final state and recovery steps; how heat behaves afterheat_domain_admin_passwordchanges, whenheat.confgets the new value but Keystone keeps the old one. - Kolla images §6, item 7: a changed
heat_domain_admin_passwordnever reaches Keystone CODE; NOT AVAILABLE whether stack-domain operations then fail authentication. - The effect on running stacks or on instances of a changed
heat_metadata_server_urlorheat_waitcondition_server_url(when the public CFN port changes) is NOT AVAILABLE: the sources only show the new value rendered intoheat.conf(roles/heat/templates/heat.conf.j2:6-7).
Service card¶
The rows of the card are explained on the Components page.
| Item | Value |
|---|---|
| Play | site.yml:724 |
| Tags | heat |
| Play host patterns | heat-api, heat-api-cfn, heat-engine |
| Node types in play (example inventory) | control |
| Node types running its containers | control |
| Containers | heat_api, heat_api_cfn, heat_engine |
deploy/reconfigure flow (tasks/) |
register → config → check-containers → clone (if) → bootstrap → flush handlers |
| Task files | bootstrap.yml, bootstrap_service.yml, check-containers.yml, check.yml, clone.yml, config.yml, config_validate.yml, copy-certs.yml, deploy-containers.yml, deploy.yml, loadbalancer.yml, main.yml, precheck.yml, pull.yml, reconfigure.yml, register.yml, stop.yml, upgrade.yml |
| Templates | heat-api-cfn.json.j2, heat-api.json.j2, heat-engine.json.j2, heat.conf.j2, wsgi-heat-api-cfn.conf.j2, wsgi-heat-api.conf.j2 |
| Handlers (definition order) | Restart heat-api container; Restart heat-api-cfn container; Restart heat-engine container |
| Key variables | enable_heat, heat_*_external_fqdn |
| Role dependencies | none |
Related¶
- Reconfigure a single service — what
-t <svc> reconfigureruns for an API service such as heat, with the risks and the verification - Keystone — a changed
heat_domain_admin_passwordreachesheat.confbut never Keystone - Stop and destroy — the
stopcommand line, whose example uses-t heat - R24 — heat-engine is one of the services named for the kill 60 s
(
docker_graceful_timeout) after SIGTERM - RabbitMQ — quorum queues on by default in 2024.1, rendered into
heat.conf - Component dependencies §6 — Heat, Horizon and Octavia come after Nova and Neutron in the deployment order
- Components — the other component pages