def wait_till_specific_host_is_in_stage(
client,
cluster_id: str,
host_name: str,
stages: List[str],
nodes_count: int = 1,
timeout: int = consts.CLUSTER_INSTALLATION_TIMEOUT / 2,
interval: int = 5,
):
log.info(f"Wait till {host_name} host is in stage {stages}")
try:
waiting.wait(
lambda: utils.are_host_progress_in_stage(
[client.get_host_by_name(cluster_id, host_name)],
stages,
nodes_count,
),
timeout_seconds=timeout,
sleep_seconds=interval,
waiting_for=f"Node to be in of the stage {stages}",
)
except BaseException:
hosts = [client.get_host_by_name(cluster_id, host_name)]
log.error(
f"All nodes stages: "
f"{[host['progress']['current_stage'] for host in hosts]} "
f"when waited for {stages}"
)
raise
def wait_till_at_least_one_host_is_in_stage(
client,
cluster_id,
stages,
nodes_count=1,
timeout=consts.CLUSTER_INSTALLATION_TIMEOUT / 2,
interval=consts.DEFAULT_CHECK_STATUSES_INTERVAL,
):
log.info(f"Wait till {nodes_count} node is in stage {stages}")
try:
waiting.wait(
lambda: utils.are_host_progress_in_stage(
client.get_cluster_hosts(cluster_id),
stages,
nodes_count,
),
timeout_seconds=timeout,
sleep_seconds=interval,
waiting_for=f"Node to be in of the stage {stages}",
)
except BaseException:
hosts = client.get_cluster_hosts(cluster_id)
log.error(
f"All nodes stages: "
f"{[host['progress']['current_stage'] for host in hosts]} "
f"when waited for {stages}"
)
raise
def download_logs(
client: InventoryClient,
cluster: dict,
dest: str,
must_gather: bool,
update_by_events: bool = False,
retry_interval: int = RETRY_INTERVAL,
):
if "hosts" not in cluster or len(cluster["hosts"]) == 0:
cluster["hosts"] = client.get_cluster_hosts(cluster_id=cluster["id"])
output_folder = get_logs_output_folder(dest, cluster)
if not is_update_needed(output_folder, update_by_events, client, cluster):
log.info(f"Skipping, no need to update {output_folder}.")
return
recreate_folder(output_folder)
recreate_folder(os.path.join(output_folder, "cluster_files"))
try:
write_metadata_file(client, cluster,
os.path.join(output_folder, "metadata.json"))
with SuppressAndLog(requests.exceptions.RequestException,
ConnectionError, KeyboardInterrupt):
client.download_metrics(os.path.join(output_folder, "metrics.txt"))
for cluster_file in (
"bootstrap.ign",
"master.ign",
"worker.ign",
"install-config.yaml",
):
with SuppressAndLog(assisted_service_client.rest.ApiException,
KeyboardInterrupt):
client.download_and_save_file(
cluster["id"], cluster_file,
os.path.join(output_folder, "cluster_files", cluster_file))
with SuppressAndLog(assisted_service_client.rest.ApiException,
KeyboardInterrupt):
download_manifests(client, cluster["id"], output_folder)
infra_env_list = set()
for host_id, infra_env_id in map(
lambda host: (host["id"], host["infra_env_id"]),
cluster["hosts"]):
with SuppressAndLog(assisted_service_client.rest.ApiException,
KeyboardInterrupt):
client.download_host_ignition(
infra_env_id, host_id,
os.path.join(output_folder, "cluster_files"))
if infra_env_id not in infra_env_list:
infra_env_list.add(infra_env_id)
with SuppressAndLog(assisted_service_client.rest.ApiException,
KeyboardInterrupt):
client.download_infraenv_events(
infra_env_id,
get_infraenv_events_path(infra_env_id, output_folder))
with SuppressAndLog(assisted_service_client.rest.ApiException,
KeyboardInterrupt):
client.download_cluster_events(
cluster["id"], get_cluster_events_path(cluster, output_folder))
shutil.copy2(
os.path.join(os.path.dirname(os.path.realpath(__file__)),
"events.html"), output_folder)
with SuppressAndLog(assisted_service_client.rest.ApiException,
KeyboardInterrupt):
are_masters_in_configuring_state = are_host_progress_in_stage(
cluster["hosts"], [HostsProgressStages.CONFIGURING], 2)
are_masters_in_join_or_done_state = are_host_progress_in_stage(
cluster["hosts"],
[HostsProgressStages.JOINED, HostsProgressStages.DONE], 2)
max_retries = MUST_GATHER_MAX_RETRIES if are_masters_in_join_or_done_state else MAX_RETRIES
is_controller_expected = cluster[
"status"] == ClusterStatus.INSTALLED or are_masters_in_configuring_state
min_number_of_logs = min_number_of_log_files(
cluster, is_controller_expected)
for i in range(max_retries):
cluster_logs_tar = os.path.join(
output_folder, f"cluster_{cluster['id']}_logs.tar")
with suppress(FileNotFoundError):
os.remove(cluster_logs_tar)
client.download_cluster_logs(cluster["id"], cluster_logs_tar)
try:
verify_logs_uploaded(
cluster_logs_tar,
min_number_of_logs,
installation_success=(
cluster["status"] == ClusterStatus.INSTALLED),
check_oc=are_masters_in_join_or_done_state,
)
break
except AssertionError as ex:
log.warning("Cluster logs verification failed: %s", ex)
# Skip sleeping on last retry
if i < MAX_RETRIES - 1:
log.info(f"Going to retry in {retry_interval} seconds")
time.sleep(retry_interval)
kubeconfig_path = os.path.join(output_folder, "kubeconfig-noingress")
with SuppressAndLog(assisted_service_client.rest.ApiException):
client.download_kubeconfig_no_ingress(cluster["id"],
kubeconfig_path)
if must_gather:
config_etc_hosts(
cluster["name"],
cluster["base_dns_domain"],
client.get_api_vip(cluster, cluster["id"]),
)
download_must_gather(kubeconfig_path, output_folder)
finally:
run_command(f"chmod -R ugo+rx '{output_folder}'")