Hey 
I am using a helm chart (influxdata/influxdb2) to install my InfluxDB, with these custom values:
image:
repository: influxdb
tag: 2.7.4-alpine
pullPolicy: IfNotPresent
service:
type: LoadBalancer
persistence:
enabled: true
storageClass: "smb-sc"
accessMode: ReadWriteOnce
size: 15Gi
mountPath: /var/lib/influxdb2
Once installed, everything gets provisioned (Persitentvolume, PVC, pods, etc) - but the pod is in a crash loop, because the database is locked during the installation.
{ │
│ "bolt-path": "/var/lib/influxdb2/influxd.bolt", │
│ "engine-path": "/var/lib/influxdb2/engine", │
│ "nats-port": 4222, │
│ "http-bind-address": ":9999" │
│ } │
│ 2024-12-04T13:16:07. info booting influxd server in the background {"system": "docker"} │
│ ts=2024-12-04T13:16:07.187954Z lvl=info msg="Welcome to InfluxDB" log_id=0tGSrj_l000 version=v2.7.4 commit=19e5c0e1b7 build_date=2023-11-08T17:07:39Z log_level=info │
│ ts=2024-12-04T13:16:07.187974Z lvl=warn msg="nats-port argument is deprecated and unused" log_id=0tGSrj_l000 │
│ ts=2024-12-04T13:16:07.198003Z lvl=info msg="Resources opened" log_id=0tGSrj_l000 service=bolt path=/var/lib/influxdb2/influxd.bolt │
│ ts=2024-12-04T13:16:07.198070Z lvl=info msg="Resources opened" log_id=0tGSrj_l000 service=sqlite path=/var/lib/influxdb2/influxd.sqlite │
│ ts=2024-12-04T13:16:07.199500Z lvl=info msg="Bringing up metadata migrations" log_id=0tGSrj_l000 service="KV migrations" migration_count=20 │
│ ts=2024-12-04T13:16:07.347685Z lvl=info msg="Bringing up metadata migrations" log_id=0tGSrj_l000 service="SQL migrations" migration_count=8 │
│ 2024-12-04T13:16:08. info pinging influxd... {"system": "docker", "ping_attempt": "0"} │
│ 2024-12-04T13:16:09. info pinging influxd... {"system": "docker", "ping_attempt": "1"} │
│ 2024-12-04T13:16:10. info pinging influxd... {"system": "docker", "ping_attempt": "2"} │
│ 2024-12-04T13:16:11. info pinging influxd... {"system": "docker", "ping_attempt": "3"} │
│ 2024-12-04T13:16:12. info pinging influxd... {"system": "docker", "ping_attempt": "4"} │
│ Error: database is locked │
│ See 'influxd -h' for help │
│ ts=2024-12-04T13:16:12.363407Z lvl=error msg="Failed to apply SQL migrations" log_id=0tGSrj_l000 error="database is locked" │
│ 2024-12-04T13:16:13. info pinging influxd... {"system": "docker", "ping_attempt": "5"} │
│ /entrypoint.sh: line 285: kill: (92) - No such process │
│ 2024-12-04T13:16:13. error influxd crashed during startup {"system": "docker", "total_pings": "6"} │
│ 2024-12-04T13:16:13. warn cleaning bolt and engine files to prevent conflicts on retry {"system": "docker", "bolt_path": "/var/lib/influxdb2/influxd.bolt", "engine_path": "/var/lib/influxdb2"}
I have found several blogpost now, describing that moving the volumeMount to a local path, rather than a network CSI (SMB, NFS, etc) it works - which i can confirm, even though the files are provisioned correctly on the SMB share.
Error after upgrade from influxdb 1.8 to 2.1 - InfluxDB 2 - InfluxData Community Forums
InfluxDB on Docker: Error: database is locked - InfluxDB 2 - InfluxData Community Forums
I have PostgreSQL, Grafana, CheckMK and similar services running on the same storageclass (smb-sc), and they are working without any issues.
My question is:
Can i skip the SQL migration scripts somehow, or even better - get InfluxDB to natively work on network attached storage?
