Skip to content

Commit 879d0f6

Browse files
authored
fix(installer): make operator readiness honest in the chart and the install gate (#3466)
## What this PR does The operator Deployment ran `replicas: 1` with `maxUnavailable: 1`, so the Deployment controller took the minimum available replica count as `1 - 1 = 0` and the `Available` condition reported `True` with no running pod at all. Measured against an operator whose image tag does not exist: the condition was `True` with `reason: MinimumReplicasAvailable` while the only pod sat in `ImagePullBackOff`, `unavailableReplicas: 1`, and no `availableReplicas` key in the status. That broke two separate things, so this fixes both. The E2E install gate asked `kubectl wait --for=condition=Available` and passed in 156 ms against that dead operator. The install then carried on and died about two minutes later on the `packages.cozystack.io` CRD wait further down the file, which sends you after CRD installation when the operator never started. It now uses `kubectl rollout status`, which reads updated and available replicas directly, and dumps the deployment, pods, operator logs and namespace events when it fails. This half is deliberately independent of the chart: `rollout status` is correct under any rollout strategy, so a later change to that strategy cannot quietly make the gate vacuous again. The chart is the other half, and fixing the gate does not cover it. `Available` is read by more than this test, and until the strategy changes an alert on that condition, a health dashboard, or a person reading `kubectl get deploy` all show green while the operator is dead. The strategy moves to `maxSurge: 1` / `maxUnavailable: 0`, which restores the ordinary meaning of the condition. Kubernetes refuses `maxSurge: 0` and `maxUnavailable: 0` together, so the surge has to move for the unavailability to reach zero. That pair is worth reading twice, because it is not a configuration I picked. At `replicas: 1` the Kubernetes default of `25%` / `25%` resolves to exactly `maxSurge: 1` and `maxUnavailable: 0`, since the surge rounds up and the unavailability rounds down. What this PR does is stop overriding the default, not introduce a tuning of its own, and the risk of the change should be judged on that basis. Moving the surge is safe here, and I checked rather than assumed, because the pair came from somewhere. It was carried over from the earlier `cozystack` Deployment, where a sidecar published host port 8123 and `hostNetwork` left the surge pod unschedulable on a single-node cluster. It got here in two steps, which is worth knowing if you go looking for the reason: by the commit before this file existed the sidecar was already gone and no port was declared, yet the pair stayed. It had outlived its own reason before it was ever copied, which is why there is no justification to find in this file's own history. On a single-node cluster the surge pod is scheduled next to the outgoing one and the rollout completes; adding a declared container port reproduces the original `didn't have free ports for the requested pod ports` failure exactly. Nothing else in the pod holds a port either: metrics and health probes are bound to `0`, and the configured webhook server never starts because the operator registers no webhook and never asks the manager for the server. The surge makes leader election load-bearing, though it covers less of the overlap than it first appears. While `maxSurge: 0` capped total pods at `replicas` a second manager process could not exist, so `--leader-elect=true` was defence in depth. With a surge pod the lease is what keeps the incoming manager's reconcile loops off the cluster while the outgoing one is still leader, and because there is no readiness probe the incoming pod counts as available as soon as its container starts, so that overlap covers the whole image pull. It does not cover everything: the operator installs CRDs, installs Flux and creates the platform PackageSource before `mgr.Start()`, and the lease is only taken inside `mgr.Start()`, so those three privileged installs run outside leader election entirely. They are safe because both pods write the same objects through server-side apply under the same field manager, which is a different guarantee from the lease. The `k8s-await-election` wrapper this operator replaced did hold the whole process until it won leadership, so if that property sounds familiar it belongs to the predecessor rather than to this binary. The sibling `opensearch-operator` chart turns leader election off at one replica for a real reason: a transient lease-renewal blip makes controller-runtime exit the process, which at one replica reads as a crashlooping Deployment. This operator installs first, when the apiserver is at its shakiest, so that scenario applies here too. I am accepting the risk knowingly, because after `maxSurge: 1` the alternative is two managers reconciling the same cluster, which is worse. The flag is pinned in the unit test so it cannot be turned off without the test going red. One consequence is permanent, so I am writing it down now instead of leaving it to be found in a year. Because any bound port breaks the surge pod under `hostNetwork`, an `httpGet` readiness probe is off the table for this Deployment for good and only an `exec` probe could ever work. So `Available` still answers "a container started", not "the operator is healthy": an operator that starts and then dies partway through installing CRDs is caught by the CRD waits below the gate, not by the gate itself. The surge also opens one new window. The operator installs CRDs and Flux before the manager starts, which puts that work outside leader election, so the incoming pod runs those applies while the outgoing pod is still leader. Both apply the same objects server-side and the outgoing pod exits moments later. `--wait` is dropped from the helm step in the same change. Helm decides a Deployment is ready with the same `replicas - maxUnavailable` arithmetic, so on the old strategy it was satisfied at zero ready pods and gated nothing, which is why the broken install got past it as well. Once the strategy is honest it would gate for real, but on its own 2m budget, and it would fail the install before the rollout gate ran and hide the diagnostics behind a bare Helm timeout. Hooks are waited for with or without the flag, and the release otherwise carries only the Deployment and its RBAC objects, so the rollout gate is the better single place to decide the operator is up. A helm unittest suite pins the strategy and all four preconditions the surge makes load-bearing: no container or initContainer declares a port, both listener arguments stay disabled, leader election stays on, and no probe uses `httpGet`. That last one follows from the two disabled listeners leaving no port for a probe to reach, and it is the worst of the four to get wrong, because the others wedge an upgrade while a `livenessProbe` against a disabled listener crashloops the pod on every cluster, fresh install included. Only the `httpGet` field is forbidden rather than the probes themselves, so an `exec` probe stays available. All four run against all three chart variants, because the template branches on variant inside the container spec and a port added under one branch would render for that variant alone. I checked the suite fails for the right reasons instead of trusting it. Reverting the strategy, declaring a `containerPort` on the operator container, declaring one under a single variant branch, declaring one on an initContainer, pointing a bind-address at a real address, turning leader election off, and adding a readiness, liveness or startup probe with `httpGet` each turns it red on its own assert. Inserting a container ahead of the operator trips the container-name pin before the argument asserts can start describing the wrong container. An `exec` probe correctly stays green. The rationale is split by audience. What an operator applying the rendered manifest has to respect stays in plain comments and travels into `_out/assets/cozystack-operator-*.yaml`: why the strategy is what it is, that no port may be declared or bound, and that leader election has to stay on. Everything that is only about this repository moves into a `{{/* */}}` template comment that stops at the chart boundary. That covers the pointer to the test and which assert pins what, where the strategy was copied from, why the webhook server on 9443 is not a third port binder, and why `opensearch-operator` reaches the opposite conclusion on leader election. The other templates in this chart already keep maintainer notes that way. ### Screenshots Not applicable, no UI changes. ### Downstream repositories I left the boxes empty on purpose rather than to skip the question, because the honest answer is not one of them and someone else should make the call. Walking the trigger map in `docs/agents/contributing.md`, nothing fires. The `hack/` triggers name `hack/e2e-prepare-cluster.bats`, `hack/package.mk`, `hack/update-crd.sh` and moves or renames under `hack/`, and this PR only edits the body of `hack/e2e-install-cozystack.bats`. The `ansible-cozystack` trigger is renaming an object the role waits on, and `deployment/cozystack-operator` keeps its name. The behaviour that role sees does change, though. It runs `kubectl wait deployment/cozystack-operator --for=condition=Available`, the same vacuous check this PR replaces here, so the chart change turns its wait from one that returns instantly into one that means something. Its `cozystack_operator_wait_timeout` defaults to 300 seconds, the same budget I gave the gate here, so a cold image pull fits and I do not think anything needs to change over there. No follow-up PR seemed warranted for a change that only makes an existing check work, but that is a judgement call rather than a trigger, which is why the boxes are empty. - [ ] No downstream repository is affected by this change - [ ] [cozystack/website](https://github.com/cozystack/website) - follow-up: - [ ] [cozystack/terraform-provider-cozystack](https://github.com/cozystack/terraform-provider-cozystack) - follow-up: - [ ] [cozystack/ansible-cozystack](https://github.com/cozystack/ansible-cozystack) - follow-up: - [ ] [cozystack/ccp](https://github.com/cozystack/ccp) - follow-up: - [ ] [cozystack/talm](https://github.com/cozystack/talm) - follow-up: - [ ] [cozystack/cozyhr](https://github.com/cozystack/cozyhr) - follow-up: - [ ] [cozystack/cozy-proxy](https://github.com/cozystack/cozy-proxy) - follow-up: - [ ] [cozystack/cozystack-telemetry-server](https://github.com/cozystack/cozystack-telemetry-server) - follow-up: - [ ] [cozystack/external-apps-example](https://github.com/cozystack/external-apps-example) - follow-up: - [ ] [cozystack/examples](https://github.com/cozystack/examples) - follow-up: ### Release note ```release-note fix(installer): the cozystack-operator Deployment no longer reports Available with zero running pods. It ran replicas: 1 with maxUnavailable: 1, which made the minimum available replica count 0, so the condition stayed True while the operator was dead and anything reading it, including the E2E install gate, saw green. The rollout strategy is now maxSurge: 1 / maxUnavailable: 0 and the install gate waits for the rollout instead of the condition. ``` <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved operator installation checks by monitoring deployment rollouts and providing deployment, pod, log, and event details when a rollout fails. * Updated operator upgrades to maintain availability during rollout, including in single-node environments. * Documented rollout requirements and behavior across supported installation configurations. * **Tests** * Added coverage for rollout settings and upgrade compatibility across supported installation variants. * Clarified Helm unit-test directory coverage documentation. <!-- end of auto-generated comment: release notes by coderabbit.ai -->
2 parents 2527e91 + 115d9bb commit 879d0f6

4 files changed

Lines changed: 297 additions & 8 deletions

File tree

‎hack/e2e-install-cozystack.bats‎

Lines changed: 33 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -117,7 +117,6 @@
117117
--namespace cozy-system \
118118
--create-namespace \
119119
--set cozystackOperator.helmReleaseInterval=30s \
120-
--wait \
121120
--timeout 2m
122121

123122
# The pre-install hook (cozy-system-labeler) must have stamped the PSA and
@@ -127,8 +126,39 @@
127126
kubectl get ns cozy-system -o jsonpath='{.metadata.labels.pod-security\.kubernetes\.io/enforce}' | grep -qx privileged
128127
kubectl get ns cozy-system -o jsonpath='{.metadata.labels.cozystack\.io/system}' | grep -qx true
129128

130-
# Verify the operator deployment is available
131-
kubectl wait deployment/cozystack-operator -n cozy-system --timeout=1m --for=condition=Available
129+
# Wait for the ROLLOUT, not for condition=Available: once maxUnavailable
130+
# reaches replicas the minimum available replica count is 0, so Available
131+
# reports True with no pod running and a broken operator image passes in
132+
# milliseconds, surfacing minutes later as a CRD timeout that blames the wrong
133+
# component. `rollout status` reads updated and available replicas directly,
134+
# and stays correct whatever strategy the chart carries.
135+
#
136+
# It still only proves a container started, not that the operator is healthy:
137+
# there is no readiness probe, so one that starts and then dies mid-install
138+
# passes here and is caught by the CRD waits below instead.
139+
#
140+
# This is the only readiness gate for the operator, since the helm step above
141+
# drops --wait, so the timeout must cover scheduling and a cold pull of an
142+
# image that is not in the prepull set.
143+
kubectl rollout status deployment/cozystack-operator -n cozy-system --timeout=5m || {
144+
echo "=== deployment/cozystack-operator rollout did not complete ==="
145+
kubectl get deployment/cozystack-operator -n cozy-system -o yaml 2>&1 || true
146+
echo "=== cozy-system pods ==="
147+
kubectl get pods -n cozy-system -o wide 2>&1 || true
148+
# The two shapes this gate exists to catch need different evidence, and
149+
# neither is in `get pods -o wide`. ImagePullBackOff produces no logs at
150+
# all, and CrashLoopBackOff puts the stack trace in the container that
151+
# already died, so describe carries the first and --previous the second.
152+
echo "=== cozystack-operator pod detail ==="
153+
kubectl describe pod -n cozy-system -l app=cozystack-operator 2>&1 || true
154+
echo "=== cozystack-operator logs ==="
155+
kubectl logs -n cozy-system -l app=cozystack-operator --all-containers --tail=50 2>&1 || true
156+
echo "=== cozystack-operator logs, previous container ==="
157+
kubectl logs -n cozy-system -l app=cozystack-operator --all-containers --previous --tail=50 2>&1 || true
158+
echo "=== cozy-system events ==="
159+
kubectl get events -n cozy-system --sort-by=.lastTimestamp 2>&1 | tail -30 || true
160+
false
161+
}
132162

133163
# Wait for operator to install CRDs (happens at startup before reconcile loop).
134164
# kubectl wait fails immediately if the CRD does not exist yet, so poll until it appears first.

‎hack/helm-unit-tests.sh‎

Lines changed: 5 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -2,9 +2,11 @@
22
set -eu
33

44
# Script to run unit tests for all Helm charts.
5-
# It iterates through directories in packages/apps, packages/extra,
6-
# packages/system, and packages/library and runs the 'test' Makefile
7-
# target if it exists.
5+
# It iterates through directories in packages/apps, packages/core,
6+
# packages/extra, packages/system, and packages/library and runs the 'test'
7+
# Makefile target if it exists. Keep this list in step with the loop below:
8+
# packages/core carries suites of its own, so dropping it to match a stale
9+
# comment would silently stop running them.
810

911
FAILED_DIRS_FILE="$(mktemp)"
1012
trap 'rm -f "$FAILED_DIRS_FILE"' EXIT

‎packages/core/installer/templates/cozystack-operator.yaml‎

Lines changed: 115 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,3 +1,86 @@
1+
{{/*
2+
Maintainer notes on the rollout strategy. The operator-facing constraints it
3+
implies are in plain # comments further down, because those ship to whoever
4+
applies the rendered _out/assets/cozystack-operator-*.yaml. Everything here
5+
is about this repository and stays inside the template.
6+
7+
tests/rollout_strategy_test.yaml pins the strategy values and the four
8+
preconditions below against every variant the chart ships, all on one shared
9+
anchor, so gating any of them on a variant is caught by the variants it was
10+
not gated to. Three preconditions are pinned airtight; precondition 2 is not,
11+
and says where the gap is rather than leaving the reader to assume coverage
12+
that is not there:
13+
14+
1. No container or initContainer declares a port. Asserted with a
15+
containers[*] wildcard rather than containers[0], because the constraint
16+
binds every container in the pod and not only the first: under hostNetwork
17+
a declared containerPort defaults to the same hostPort, and the surge pod
18+
then has nowhere to go on a single-node cluster.
19+
2. Neither listener is bound at runtime. --metrics-bind-address=0 and
20+
--health-probe-bind-address=0 are asserted explicitly, because the
21+
operator's own defaults are :8080 and :8081. The arguments are the only
22+
thing holding this, and the scheduler cannot see it either way. This is
23+
the loosely pinned one, and the reason is worth stating precisely so
24+
nobody trusts the wrong model of it. contains is a membership test with
25+
no notion of position, while Go's flag package keeps the last occurrence,
26+
so a duplicate appearing AFTER the disabled one wins and the assert still
27+
passes; a duplicate before it loses, and the assert passes just the same.
28+
Moving the disabled flag last therefore fixes the process but not the
29+
test, which stays green either way. notContains on :8080 and :8081 closes
30+
that for those two addresses, again regardless of position, and they are
31+
the ones a values-gated block added to let users enable metrics would
32+
reach for, since they are this binary's own flag defaults. Any other
33+
address slips through: helm-unittest cannot express "this flag appears no
34+
more than once". So changing a value in place is caught, and duplicating
35+
one is caught only for those two addresses.
36+
3. --leader-elect stays on. maxSurge: 1 means a second manager process
37+
exists by construction on every upgrade, so the lease is what keeps the
38+
incoming pod's reconcile loops off the cluster until the outgoing one is
39+
gone. Turning it off lets two managers reconcile at once.
40+
4. No probe uses httpGet. This follows from 2: with both listeners disabled
41+
an httpGet probe has no port left to target. Only the httpGet field is
42+
asserted absent, not the probes themselves, so an exec probe stays
43+
legitimate — but under maxUnavailable: 0 an exec probe that cannot pass
44+
while the outgoing pod is still alive wedges the rollout permanently
45+
rather than transiently, since the outgoing pod is not retired until the
46+
incoming one is available. A readiness gate that depends on the old
47+
process being gone deadlocks against the strategy.
48+
49+
Read precondition 3 for exactly what it says and no more. The lease governs
50+
the reconcile loops; it does not govern everything the two pods do at once,
51+
and two overlaps sit outside it. The routine one happens on every healthy
52+
upgrade and is bounded by the outgoing pod's termination grace: for that
53+
moment an old binary keeps reconciling against schema the incoming pod has
54+
already moved. The rare one needs the outgoing pod to be broken and
55+
restarting, and it is the worse of the two, because one ordering of it
56+
silently reverts the incoming pod's install phase and does not heal itself
57+
afterwards. Both are known and accepted, not oversights. Unlike the binders
58+
below, neither is hypothetical.
59+
60+
What pays for carrying them, since Recreate would make Available honest with
61+
no second pod at all: when the replacement never starts — a surge pod that
62+
cannot be scheduled, an image that cannot be pulled — maxUnavailable: 0 keeps
63+
the outgoing operator running while the rollout wedges, where Recreate would
64+
have deleted it first and left none. That is the whole of the gain, and it
65+
covers only that class. A replacement that starts and then dies is already
66+
counted available, so the outgoing pod is gone by then under either strategy.
67+
68+
The variant loop is not ceremony: the template branches on variant inside the
69+
container spec, so a port added under one branch renders for that variant
70+
alone and a single-variant suite would not see it. The container-name assert
71+
pins containers[0] for the three argument asserts, which would otherwise start
72+
describing a different container if one were inserted ahead of the operator.
73+
74+
The manager is also configured with a webhook server on 9443, which reads
75+
like a third binder and is not one: controller-runtime adds that server to
76+
its runnables only inside GetWebhookServer(), and this binary registers no
77+
webhook and never calls it, so nothing listens. Registering one would bind
78+
9443 in the host namespace and wedge the rollout the same way, and no
79+
chart-level test can reach that — it is a Go-side change. PprofBindAddress is
80+
a fourth binder of the same kind: unset today, with no flag exposing it and
81+
no path to it from this chart, so it is inert for the same reason and would
82+
wedge the rollout for the same reason if it were ever set.
83+
*/}}
184
{{- $validVariants := list "talos" "generic" "hosted" -}}
285
{{- if not (has .Values.cozystackOperator.variant $validVariants) -}}
386
{{- fail (printf "Invalid cozystackOperator.variant %q: must be one of talos, generic, hosted" .Values.cozystackOperator.variant) -}}
@@ -35,8 +118,34 @@ spec:
35118
strategy:
36119
type: RollingUpdate
37120
rollingUpdate:
38-
maxSurge: 0
39-
maxUnavailable: 1
121+
# This pair is what Kubernetes itself would pick here: the default
122+
# 25% / 25% resolves at replicas: 1 to maxSurge 1 and maxUnavailable 0,
123+
# because the surge rounds up and the unavailability rounds down. Spelling
124+
# it out is not tuning, it is declining to override the default.
125+
# maxUnavailable must stay 0. At replicas: 1 anything higher makes the
126+
# minimum available replica count replicas - maxUnavailable = 0, and the
127+
# Available condition then reports True with no running pod at all, which
128+
# hides a dead operator from every consumer of that status.
129+
# Kubernetes rejects maxUnavailable: 0 together with maxSurge: 0, so the
130+
# surge cannot go the other way either. The surge pod shares the node
131+
# network namespace (hostNetwork below), so nothing in this pod may need
132+
# a port, and the two ways of needing one fail differently. Declaring a
133+
# containerPort makes the surge pod unschedulable on a single-node
134+
# cluster: it never runs, so maxUnavailable: 0 keeps the outgoing pod
135+
# serving and the rollout simply wedges. Binding a port at runtime, by
136+
# pointing --metrics-bind-address or --health-probe-bind-address at a
137+
# real address, is worse, because this pod has no readiness probe and
138+
# counts as available the moment its container starts. The outgoing pod
139+
# is retired at that point, and the incoming one dies with EADDRINUSE
140+
# afterwards, leaving no operator at all rather than the old one. Do not
141+
# read maxUnavailable: 0 as protection against that second class; what
142+
# catches it is whatever waits on the resources the operator is supposed
143+
# to create.
144+
# This also rules out an httpGet probe of any kind on this Deployment: the
145+
# port it would target is one of the two disabled above. Only an exec
146+
# probe can ever work here.
147+
maxSurge: 1
148+
maxUnavailable: 0
40149
template:
41150
metadata:
42151
labels:
@@ -47,6 +156,10 @@ spec:
47156
- name: cozystack-operator
48157
image: "{{ .Values.cozystackOperator.image }}"
49158
args:
159+
# Must stay on. maxSurge: 1 above means a second manager pod exists on
160+
# every upgrade, and the lease is what keeps the incoming one's
161+
# reconcile loops off the cluster until the outgoing one is gone.
162+
# Turning it off lets two managers reconcile the same cluster at once.
50163
- --leader-elect=true
51164
- --install-flux=true
52165
# The operator applies embedded CRDs via server-side apply on every
Lines changed: 144 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,144 @@
1+
suite: cozystack-operator rollout strategy
2+
3+
# Pins the rollout strategy and the four preconditions the surge pod makes
4+
# load-bearing: no declared port, no bound listener, leader election on, no
5+
# httpGet probe. Each is checked against every variant the chart ships. The
6+
# bound-listener one is pinned only partially, for a reason recorded next to
7+
# the asserts themselves; the other three are airtight.
8+
#
9+
# Why the strategy is what it is, why each precondition matters and how each one
10+
# fails: the maintainer note at the top of templates/cozystack-operator.yaml.
11+
# That note is the single owner of the reasoning; this file only asserts it.
12+
13+
templates:
14+
- templates/cozystack-operator.yaml
15+
16+
tests:
17+
- it: keeps the strategy and every surge precondition on the talos variant
18+
set:
19+
cozystackOperator.variant: talos
20+
asserts: &surgeSafe
21+
# The strategy rides the same anchor as the preconditions rather than
22+
# sitting in a talos-only case of its own. It renders outside every
23+
# variant branch today, so one case would pass, but that is a property
24+
# of the template rather than a guarantee: move the block inside a
25+
# branch and a single-variant assert stops seeing the other two.
26+
- equal:
27+
path: spec.replicas
28+
value: 1
29+
documentSelector:
30+
path: kind
31+
value: Deployment
32+
# Pinned alongside the two values below because they are meaningless
33+
# without it: MaxUnavailable() reports 0 for any strategy that is not
34+
# RollingUpdate, so a switch to Recreate would satisfy the intent of this
35+
# test by a different route while leaving the rollingUpdate block behind
36+
# as dead configuration the API server then rejects at apply.
37+
- equal:
38+
path: spec.strategy.type
39+
value: RollingUpdate
40+
documentSelector:
41+
path: kind
42+
value: Deployment
43+
- equal:
44+
path: spec.strategy.rollingUpdate.maxUnavailable
45+
value: 0
46+
documentSelector:
47+
path: kind
48+
value: Deployment
49+
- equal:
50+
path: spec.strategy.rollingUpdate.maxSurge
51+
value: 1
52+
documentSelector:
53+
path: kind
54+
value: Deployment
55+
- equal:
56+
path: spec.template.spec.hostNetwork
57+
value: true
58+
documentSelector:
59+
path: kind
60+
value: Deployment
61+
- notExists:
62+
path: spec.template.spec.containers[*].ports
63+
documentSelector:
64+
path: kind
65+
value: Deployment
66+
- notExists:
67+
path: spec.template.spec.initContainers[*].ports
68+
documentSelector:
69+
path: kind
70+
value: Deployment
71+
# The three argument asserts read containers[0], so pin which container
72+
# that is. A container inserted ahead of the operator would otherwise
73+
# shift the index and leave them asserting about something else.
74+
- equal:
75+
path: spec.template.spec.containers[0].name
76+
value: cozystack-operator
77+
documentSelector:
78+
path: kind
79+
value: Deployment
80+
# contains is a membership test with no notion of position, while Go's
81+
# flag package keeps the last occurrence, so these two alone stay green
82+
# whether a duplicate lands before the =0 and loses or after it and binds
83+
# the port — they cannot tell the two apart. Moving the =0 flag last
84+
# would fix the process without making these asserts catch anything. The
85+
# notContains pair below closes that for the two addresses this binary
86+
# defaults its own flags to; any other address still slips through.
87+
- contains:
88+
path: spec.template.spec.containers[0].args
89+
content: --metrics-bind-address=0
90+
documentSelector:
91+
path: kind
92+
value: Deployment
93+
- notContains:
94+
path: spec.template.spec.containers[0].args
95+
content: --metrics-bind-address=:8080
96+
documentSelector:
97+
path: kind
98+
value: Deployment
99+
- contains:
100+
path: spec.template.spec.containers[0].args
101+
content: --health-probe-bind-address=0
102+
documentSelector:
103+
path: kind
104+
value: Deployment
105+
- notContains:
106+
path: spec.template.spec.containers[0].args
107+
content: --health-probe-bind-address=:8081
108+
documentSelector:
109+
path: kind
110+
value: Deployment
111+
- contains:
112+
path: spec.template.spec.containers[0].args
113+
content: --leader-elect=true
114+
documentSelector:
115+
path: kind
116+
value: Deployment
117+
# Only the httpGet field is asserted absent, not the probes themselves, so
118+
# an exec probe stays legitimate.
119+
- notExists:
120+
path: spec.template.spec.containers[*].readinessProbe.httpGet
121+
documentSelector:
122+
path: kind
123+
value: Deployment
124+
- notExists:
125+
path: spec.template.spec.containers[*].livenessProbe.httpGet
126+
documentSelector:
127+
path: kind
128+
value: Deployment
129+
- notExists:
130+
path: spec.template.spec.containers[*].startupProbe.httpGet
131+
documentSelector:
132+
path: kind
133+
value: Deployment
134+
135+
- it: keeps the strategy and every surge precondition on the generic variant
136+
set:
137+
cozystackOperator.variant: generic
138+
cozystack.apiServerHost: api.example.com
139+
asserts: *surgeSafe
140+
141+
- it: keeps the strategy and every surge precondition on the hosted variant
142+
set:
143+
cozystackOperator.variant: hosted
144+
asserts: *surgeSafe

0 commit comments

Comments
 (0)