[antithesis] Fix broken flag handling and improve image testing (#3299)

.golangci.yml

@@ -170,16 +170,10 @@ linters-settings:
         disabled: false
       # https://github.com/mgechev/revive/blob/master/RULES_DESCRIPTIONS.md#unhandled-error
       - name: unhandled-error
-        disabled: false
-        arguments:
-          - "fmt\\.Fprint"
-          - "fmt\\.Fprintf"
-          - "fmt\\.Fprintln"
-          - "fmt\\.Print"
-          - "fmt\\.Printf"
-          - "fmt\\.Println"
-          - "math/rand\\.Read"
-          - "strings\\.Builder\\.WriteString"
+        # prefer the errcheck linter since it can be disabled directly with nolint directive
+        # but revive's disable directive (e.g. //revive:disable:unhandled-error) is not
+        # supported when run under golangci_lint
+        disabled: true
       # https://github.com/mgechev/revive/blob/master/RULES_DESCRIPTIONS.md#unused-parameter
       - name: unused-parameter
         disabled: false

scripts/lib_test_antithesis_images.sh

@@ -46,13 +46,17 @@ docker cp "${CONTAINER_NAME}":/docker-compose.yml "${COMPOSE_FILE}"
 # Copy the volume paths out of the container
 docker cp "${CONTAINER_NAME}":/volumes "${TMPDIR}/"
 
-# Run the docker compose project for 30 seconds without error. Local
-# network bootstrap is ~6s, but github workers can be much slower.
-${COMPOSE_CMD} up -d
-sleep 30
-if ${COMPOSE_CMD} ps -q | xargs docker inspect -f '{{ .State.Status }}' | grep -v 'running'; then
-  echo "An error occurred."
+# Wait for up to TIMEOUT for the workload to emit HEALTHY_MESSAGE to indicate that all nodes are
+# reporting healthy. This indicates that the workload has been correctly configured. Subsequent
+# validation will need to be tailored to a given workload implementation.
+
+TIMEOUT=30s
+HEALTHY_MESSAGE="all nodes reported healthy"
+
+if timeout "${TIMEOUT}" bash -c "${COMPOSE_CMD} up 2>&1 | grep -m 1 '${HEALTHY_MESSAGE}'"; then
+  echo "Saw log containing '${HEALTHY_MESSAGE}'"
+  echo "Successfully invoked the antithesis test setup configured by ${IMAGE_NAME}:${IMAGE_TAG}"
+else
+  echo "Failed to see log containing '${HEALTHY_MESSAGE}' within ${TIMEOUT}"
   exit 1
 fi
-
-echo "Successfully invoked the antithesis test setup configured by ${IMAGE_NAME}:${IMAGE_TAG}"

tests/antithesis/compose.go

@@ -9,7 +9,6 @@ import (
 	"os"
 	"path/filepath"
 	"strconv"
-	"strings"
 
 	"github.com/compose-spec/compose-go/types"
 	"gopkg.in/yaml.v3"
@@ -129,7 +128,7 @@ func newComposeProject(network *tmpnet.Network, nodeImageName string, workloadIm
 	baseNetworkAddress := "10.0.20"
 
 	services := make(types.Services, len(network.Nodes)+1)
-	uris := make([]string, len(network.Nodes))
+	uris := make(CSV, len(network.Nodes))
 	var (
 		bootstrapIP  string
 		bootstrapIDs string
@@ -230,16 +229,16 @@ func newComposeProject(network *tmpnet.Network, nodeImageName string, workloadIm
 	}
 
 	workloadEnv := types.Mapping{
-		envVarName(URIsKey): strings.Join(uris, " "),
+		envVarName(EnvPrefix, URIsKey): uris.String(),
 	}
-	chainIDs := []string{}
+	chainIDs := CSV{}
 	for _, subnet := range network.Subnets {
 		for _, chain := range subnet.Chains {
 			chainIDs = append(chainIDs, chain.ChainID.String())
 		}
 	}
 	if len(chainIDs) > 0 {
-		workloadEnv[envVarName(ChainIDsKey)] = strings.Join(chainIDs, " ")
+		workloadEnv[envVarName(EnvPrefix, ChainIDsKey)] = chainIDs.String()
 	}
 
 	workloadName := "workload"
@@ -277,8 +276,7 @@ func newComposeProject(network *tmpnet.Network, nodeImageName string, workloadIm
 func keyMapToEnvVarMap(keyMap types.Mapping) types.Mapping {
 	envVarMap := make(types.Mapping, len(keyMap))
 	for key, val := range keyMap {
-		// e.g. network-id -> AVAGO_NETWORK_ID
-		envVar := strings.ToUpper(config.EnvPrefix + "_" + config.DashesToUnderscores.Replace(key))
+		envVar := envVarName(config.EnvPrefix, key)
 		envVarMap[envVar] = val
 	}
 	return envVarMap

tests/antithesis/config.go

@@ -11,6 +11,7 @@ import (
 
 	"github.com/stretchr/testify/require"
 
+	"github.com/ava-labs/avalanchego/config"
 	"github.com/ava-labs/avalanchego/tests"
 	"github.com/ava-labs/avalanchego/tests/fixture/e2e"
 	"github.com/ava-labs/avalanchego/tests/fixture/tmpnet"
@@ -54,15 +55,15 @@ func NewConfigWithSubnets(tc tests.TestContext, defaultNetwork *tmpnet.Network,
 	flag.Parse()
 
 	// Env vars take priority over flags
-	envURIs := os.Getenv(envVarName(URIsKey))
+	envURIs := os.Getenv(envVarName(EnvPrefix, URIsKey))
 	if len(envURIs) > 0 {
-		// CSV.Set doesn't actually return an error
-		_ = uris.Set(envURIs)
+		//nolint:errcheck // CSV.Set doesn't actually return an error
+		uris.Set(envURIs)
 	}
-	envChainIDs := os.Getenv(envVarName(ChainIDsKey))
+	envChainIDs := os.Getenv(envVarName(EnvPrefix, ChainIDsKey))
 	if len(envChainIDs) > 0 {
-		// CSV.Set doesn't actually return an error
-		_ = uris.Set(envChainIDs)
+		//nolint:errcheck // CSV.Set doesn't actually return an error
+		chainIDs.Set(envChainIDs)
 	}
 
 	// Use the network configuration provided
@@ -126,6 +127,7 @@ func (c *CSV) Set(value string) error {
 	return nil
 }
 
-func envVarName(key string) string {
-	return strings.ToUpper(EnvPrefix + "_" + key)
+func envVarName(prefix string, key string) string {
+	// e.g. MY_PREFIX, network-id -> MY_PREFIX_NETWORK_ID
+	return strings.ToUpper(prefix + "_" + config.DashesToUnderscores.Replace(key))
 }