Skip to content

scylla_image_setup: avoid script error when perftune.py failed #787

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Oct 23, 2025
Merged

scylla_image_setup: avoid script error when perftune.py failed #787

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
00cd3cb
scylla_image_setup: switch from print() to logging
syuu1228 Sep 1, 2025
ba96671
scylla_image_setup: avoid script error when perftune.py failed
syuu1228 Sep 1, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
34 changes: 31 additions & 3 deletions common/scylla_image_setup
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,11 +6,27 @@

import os
import sys
import logging
import subprocess
import re
from pathlib import Path
from lib.scylla_cloud import get_cloud_instance, is_gce, is_azure, is_redhat_variant
from lib.log import setup_logging
from lib.scylla_cloud import get_cloud_instance, is_ec2, is_gce, is_azure, is_redhat_variant
from subprocess import run

LOGGER = logging.getLogger(__name__)

def disable_perftune():
# There is no way to disable SET_CLOCKSOURCE from scylla_sysconfig_setup
# once it enabled, so we have to directly editing the sysconfig file
with open('/etc/default/scylla-server') as f:
sysconfig = f.read()
sysconfig = re.sub(r'^SET_CLOCKSOURCE=.*$', 'SET_CLOCKSOURCE=no', sysconfig, flags=re.MULTILINE)
with open('/etc/default/scylla-server', 'w') as f:
f.write(sysconfig)

if __name__ == '__main__':
setup_logging()
if is_azure():
swap_directory = Path('/mnt')
swap_unit = Path('/etc/systemd/system/mnt-swapfile.swap')
Expand All @@ -27,7 +43,19 @@ if __name__ == '__main__':
cloud_instance = get_cloud_instance()
run('/opt/scylladb/scylla-machine-image/scylla_configure.py', shell=True, check=True)

run('/opt/scylladb/scripts/scylla_sysconfig_setup --nic eth0 --setup-nic', shell=True, check=True)
try:
run('/opt/scylladb/scripts/scylla_sysconfig_setup --nic eth0 --setup-nic', shell=True, check=True)
except subprocess.CalledProcessError as e:
# There are kernel bug on some aarch64 instance types on EC2, it does not provide NUMA topology
# information, perftune crashes because of that.
# Only possible workaround is disable perftune for these instance types, for now.
# see: scylladb/seastar#2925
affected_instance_types = ['im4gn.8xlarge', 'im4gn.16xlarge']
if e.returncode == 3 and is_ec2() and cloud_instance.instancetype in affected_instance_types:
disable_perftune()
LOGGER.warning('Failed to enable perftune.py, continue without using it')
else:
raise e
if os.path.ismount('/var/lib/scylla'):
if cloud_instance.is_supported_instance_class():
# We run io_setup only when ehpemeral disks are available
Expand All @@ -44,5 +72,5 @@ if __name__ == '__main__':
run('systemctl is-active -q fstrim.timer && systemctl disable fstrim.timer', shell=True, check=True)

if not os.path.ismount('/var/lib/scylla') and not cloud_instance.is_dev_instance_type():
print('Failed to initialize RAID volume!')
LOGGER.error('Failed to initialize RAID volume!')
machine_image_configured.touch()