Troubleshooting¶
Fixes for the problems you are most likely to hit with the Vector Log Pipeline.
HTTPS ingest returns 401¶
Cause: the request has no credentials, or the wrong ones.
Fix: send HTTP basic authentication with user vector and the password
from /var/lib/vector/vector.env. If you changed the password, check that
vector.env and the http_in password in /etc/vector/vector.yaml match.
A file in /var/lib/vector/in/ is not picked up¶
Cause: the file name does not end in .log, or the vector user cannot
read it.
Fix: use a .log name and make the file readable, for example
chmod 0644 /var/lib/vector/in/<file>.log.
A record is missing from the output¶
Cause: its level is debug, so the redaction step drops it on purpose.
Fix: change the record's level, or change the redact transform in
/etc/vector/vector.yaml.
Vector does not start after a config change¶
Cause: the configuration does not validate.
Fix: run the validation yourself to see the error:
Fix the file and restart with sudo systemctl restart vector.
An S3 sink fails its health check¶
Cause: usually a missing IAM role, a policy that does not allow
s3:ListBucket on the bucket, or a wrong bucket or region.
Fix: compare your role with
/opt/sushi/etc/vector/iam-s3-sink.json.example, check the bucket and region
in the sink, and read sudo journalctl -u vector -n 50 --no-pager.
First boot failed¶
Cause: sushi-vector-firstboot.service failed, so Vector did not start.
Fix: look at its log, then run it again and start Vector:
sudo journalctl -u sushi-vector-firstboot.service --no-pager
sudo systemctl restart sushi-vector-firstboot
sudo systemctl start vector
If it still fails, email info@softwaresushi.com with the instance ID and the log output.