mirror of
https://github.com/mulgadc/spinifex.git
synced 2026-10-20 03:11:46 +00:00
Josh Siviour
92fa467c90
* refactor(e2e/multinode): flatten phase 11 + parallelize bucket #1 + drop phase numbers Split monolithic phase11_PlacementAndNATGateway (~430 lines, 8 inline sub-steps) into 6 top-level Tests, each owning its own VPC graph via t.Cleanup and layered setup helpers (spreadSetupVPC, spreadSetupBastion, spreadSetupPrivateTrio, spreadSetupNATGW). The cleanup-ordering test flips *bool latches as each explicit teardown step succeeds so the LIFO unwind is a no-op on the success path. Add t.Parallel() to read-only / independent Tests: ClusterHealth, InstanceDistribution, GuestSSH, VPCNetworking. Deviates from the mulga-siv-127 bead's bucket-#1 listing by keeping CrossNodeGateway sequential (asserts a snapshot equality across gateways that the parallel VPC test would break by churning instance count mid-assert) and CrossNodeOps sequential (stops/starts trio[0], would race GuestSSH). Both deviations documented in tests_test.go + plan doc. Drop the bash phase numbering (phaseN_X funcs + "Phase N" log strings) from the rest of the suite — 10 funcs renamed to runX. Phase numbers were a port-time crutch; the Test names now describe each scenario and phase numbers ageing into noise as bash is retired in Stage H. * refactor(e2e/multinode): collapse spread Tests + control execution order Two changes wedged together since they touch the same files: 1. Collapse the 6 top-level TestMultinodeSpread* Tests into a single TestMultinodeSpread with 6 t.Run sub-tests sharing one VPC + bastion + private trio + NAT GW setup chain. The 6-test split paid setup cost 6× — bastion launch + private trio launch + NAT GW provision per Test pushed phase-11 wall-clock near 60min vs ~15min monolithic. Shared- setup layout keeps JUnit granularity (one row per sub-test) while restoring the monolithic budget. Setup helpers register t.Cleanup against the outer test so teardown runs once at the end (LIFO: NAT GW → trio → bastion → VPC). 2. Move the TestMultinodeSpread wrapper into tests_test.go so file-source order controls execution position. Previously the func lived in placement_nat_test.go which sorts before preflight_test.go and tests_test.go alphabetically — Spread ran FIRST in the sequential bucket, defeating Preflight's fail-fast on a sick cluster. Wrapper now runs after NodeRecovery so the cluster has been fully degrade-tested before Spread's 4-VM + NAT GW + custom-VPC graph spins up.
52 lines
1.9 KiB
Go
52 lines
1.9 KiB
Go
//go:build e2e
|
|
|
|
package multinode
|
|
|
|
import (
|
|
"testing"
|
|
|
|
"github.com/aws/aws-sdk-go/service/ec2"
|
|
"github.com/mulgadc/spinifex/tests/e2e/harness"
|
|
"github.com/stretchr/testify/require"
|
|
)
|
|
|
|
// runCrossNodeGateway is the Go port of cross-node gateway access
|
|
// (run-multinode-e2e.sh:828-851). Drives DescribeInstances through every
|
|
// node's gateway and asserts each returns the same instance count as node1
|
|
// (the test runner's local gateway).
|
|
//
|
|
// Catches the regression where the daemon answers only locally-hosted
|
|
// instances instead of fanning out via NATS — i.e. a stale awsgw routing
|
|
// table or a queue-group misconfiguration.
|
|
func runCrossNodeGateway(t *testing.T, fix *Fixture) {
|
|
harness.Phase(t, "Multinode — Cross-Node Gateway Access")
|
|
|
|
// Trigger trio launch so the baseline count is meaningful.
|
|
_ = needInstanceTrio(t, fix)
|
|
|
|
baseline, err := fix.AWS.EC2.DescribeInstances(&ec2.DescribeInstancesInput{})
|
|
require.NoError(t, err, "baseline describe-instances")
|
|
want := countInstances(baseline)
|
|
harness.Detail(t, "baseline_node", fix.Cluster.Nodes[0].Name, "instances", want)
|
|
require.Greater(t, want, 0, "baseline returned 0 instances (trio not registered?)")
|
|
|
|
// Bash skips node1 (the local one) — its baseline already represents that gateway.
|
|
for _, n := range fix.Cluster.Nodes[1:] {
|
|
harness.Step(t, "DescribeInstances via %s (%s)", n.Name, n.Addr)
|
|
cli := harness.AWSClientForGateway(t, fix.Env, n)
|
|
out, err := cli.EC2.DescribeInstances(&ec2.DescribeInstancesInput{})
|
|
require.NoErrorf(t, err, "describe-instances via %s", n.Name)
|
|
got := countInstances(out)
|
|
harness.Detail(t, "node", n.Name, "instances", got)
|
|
require.Equalf(t, want, got, "%s returned %d instances, baseline %d (cross-node fan-out broken)", n.Name, got, want)
|
|
}
|
|
}
|
|
|
|
func countInstances(out *ec2.DescribeInstancesOutput) int {
|
|
n := 0
|
|
for _, r := range out.Reservations {
|
|
n += len(r.Instances)
|
|
}
|
|
return n
|
|
}
|