vitess-operator
vitess-operator copied to clipboard
When configured backup, backup parameters not added to vtcld pods
When configured backup, backup parameters not added to vtcld pods Product version Vitess version 14.0.1 Operator version 2.7.2
Problem backup arguments are not added by operator to vtcld pods it causes that is not possible to do list of backups with vtctlclient
VitessCluster Configuration
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
labels:
app: ope-vitess-backup
name: ope-vitess-backup
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 10Gi
# storageClassName: sharedfs
---
apiVersion: planetscale.com/v2
kind: VitessCluster
metadata:
name: ope-vitess-cluster
spec:
backup:
engine: xtrabackup
locations:
- volume:
persistentVolumeClaim:
claimName: ope-vitess-backup
cells:
- gateway:
extraFlags:
mysql_auth_server_impl: none
mysql_server_version: 8.0.23-Vitess
replicas: 2
resources:
limits:
memory: 256Mi
requests:
cpu: 200m
memory: 256Mi
name: region1
globalLockserver:
etcd:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- podAffinityTerm:
labelSelector:
matchExpressions:
- key: planetscale.com/component
operator: In
values:
- vttablet
topologyKey: kubernetes.io/hostname
weight: 100
dataVolumeClaimTemplate:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 2Gi
# storageClassName: sharedfs
resources:
limits:
cpu: 100m
memory: 128Mi
requests:
cpu: 100m
memory: 128Mi
images:
mysqld:
mysql80Compatible: docker.io/vitess/lite:v14.0.1-mysql80
mysqldExporter: docker.io/prom/mysqld-exporter:v0.11.0
vtbackup: docker.io/vitess/lite:v14.0.1-mysql80
vtctld: docker.io/vitess/lite:v14.0.1-mysql80
vtgate: docker.io/vitess/lite:v14.0.1-mysql80
vtorc: docker.io/vitess/lite:v14.0.1-mysql80
vttablet: docker.io/vitess/lite:v14.0.1-mysql80
keyspaces:
- databaseName: ope
durabilityPolicy: semi_sync
name: ope
partitionings:
- equal:
parts: 1
shardTemplate:
databaseInitScriptSecret:
key: init_db.sql
name: ope-vitess-cluster-config
replication:
enforceSemiSync: true
recoverRestartedMaster: false
tabletPools:
- cell: region1
dataVolumeClaimTemplate:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 10Gi
# storageClassName: local-path
mysqld:
configOverrides: |
[mysqld]
lower_case_table_names = 1
resources:
limits:
cpu: 200m
memory: 1Gi
requests:
cpu: 200m
memory: 1Gi
replicas: 3
type: replica
vttablet:
extraFlags:
db_charset: utf8mb4
disable_active_reparents: "true"
queryserver-config-transaction-timeout: "300"
resources:
limits:
cpu: 200m
memory: 1Gi
requests:
cpu: 200m
memory: 1Gi
vitessOrchestrator:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- podAffinityTerm:
labelSelector:
matchExpressions:
- key: planetscale.com/component
operator: In
values:
- vttablet
topologyKey: kubernetes.io/hostname
weight: 100
configSecret:
key: orc_config.json
name: ope-vitess-cluster-config
resources:
limits:
memory: 128Mi
requests:
cpu: 100m
memory: 128Mi
vitessDashboard:
cells:
- region1
extraFlags:
security_policy: read-only
replicas: 2
resources:
limits:
memory: 128Mi
requests:
cpu: 200m
memory: 128Mi
Command
alias vc='kubectl exec -i -t $(kubectl get pod -l "planetscale.com/component=vtctld" -o jsonpath="{.items[0].metadata.name}") -- /vt/bin/vtctlclient --server $(kubectl get svc -l "planetscale.com/component=vtctld" -o jsonpath="{.items[0].metadata.name}"):15999'
vc ListBackups ope/-
Result
ListBackups Error: rpc error: code = Unknown desc = no registered implementation of BackupStorage
E0805 13:55:55.827184 67 main.go:103] remote error: rpc error: code = Unknown desc = no registered implementation of BackupStorage
command terminated with exit code 1
I was able to configure vtcld pods with missing backup parameters and with add extraVolume for correct work retrieve list of backups, I have added the same backup params that were added to vttablets pods
Parameters that I have added
backup_engine_implementation: xtrabackup
backup_storage_compress: "true"
backup_storage_implementation: file
file_backup_storage_root: /vt/backups/ope-vitess-cluster
xbstream_restore_flags: --parallel=3
xtrabackup_backup_flags: --parallel=1
xtrabackup_stream_mode: xbstream
xtrabackup_stripes: "8"
xtrabackup_user: vt_dba
But I think that backup parameters should be added to vtcld pods by vitess operator, and only one point of backup configuration should be in backup section https://github.com/planetscale/vitess-operator/blob/main/docs/api.md#planetscale.com/v2.ClusterBackupSpec Could you please fix this issue
Workaround with add parameters and exraVolume
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
labels:
app: ope-vitess-backup
name: ope-vitess-backup
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 10Gi
# storageClassName: sharedfs
---
apiVersion: planetscale.com/v2
kind: VitessCluster
metadata:
name: ope-vitess-cluster
spec:
backup:
engine: xtrabackup
locations:
- volume:
persistentVolumeClaim:
claimName: ope-vitess-backup
cells:
- gateway:
extraFlags:
mysql_auth_server_impl: none
mysql_server_version: 8.0.23-Vitess
replicas: 2
resources:
limits:
memory: 256Mi
requests:
cpu: 200m
memory: 256Mi
name: region1
globalLockserver:
etcd:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- podAffinityTerm:
labelSelector:
matchExpressions:
- key: planetscale.com/component
operator: In
values:
- vttablet
topologyKey: kubernetes.io/hostname
weight: 100
dataVolumeClaimTemplate:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 2Gi
# storageClassName: sharedfs
resources:
limits:
cpu: 100m
memory: 128Mi
requests:
cpu: 100m
memory: 128Mi
images:
mysqld:
mysql80Compatible: docker.io/vitess/lite:v14.0.1-mysql80
mysqldExporter: docker.io/prom/mysqld-exporter:v0.11.0
vtbackup: docker.io/vitess/lite:v14.0.1-mysql80
vtctld: docker.io/vitess/lite:v14.0.1-mysql80
vtgate: docker.io/vitess/lite:v14.0.1-mysql80
vtorc: docker.io/vitess/lite:v14.0.1-mysql80
vttablet: docker.io/vitess/lite:v14.0.1-mysql80
keyspaces:
- databaseName: ope
durabilityPolicy: semi_sync
name: ope
partitionings:
- equal:
parts: 1
shardTemplate:
databaseInitScriptSecret:
key: init_db.sql
name: ope-vitess-cluster-config
replication:
enforceSemiSync: true
recoverRestartedMaster: false
tabletPools:
- cell: region1
dataVolumeClaimTemplate:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 10Gi
# storageClassName: local-path
mysqld:
configOverrides: |
[mysqld]
lower_case_table_names = 1
resources:
limits:
cpu: 200m
memory: 1Gi
requests:
cpu: 200m
memory: 1Gi
replicas: 3
type: replica
vttablet:
extraFlags:
db_charset: utf8mb4
disable_active_reparents: "true"
queryserver-config-transaction-timeout: "300"
resources:
limits:
cpu: 200m
memory: 1Gi
requests:
cpu: 200m
memory: 1Gi
vitessOrchestrator:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- podAffinityTerm:
labelSelector:
matchExpressions:
- key: planetscale.com/component
operator: In
values:
- vttablet
topologyKey: kubernetes.io/hostname
weight: 100
configSecret:
key: orc_config.json
name: ope-vitess-cluster-config
resources:
limits:
memory: 128Mi
requests:
cpu: 100m
memory: 128Mi
vitessDashboard:
cells:
- region1
extraFlags:
backup_engine_implementation: xtrabackup
backup_storage_compress: "true"
backup_storage_implementation: file
file_backup_storage_root: /vt/backups/ope-vitess-cluster
security_policy: read-only
xbstream_restore_flags: --parallel=3
xtrabackup_backup_flags: --parallel=1
xtrabackup_stream_mode: xbstream
xtrabackup_stripes: "8"
xtrabackup_user: vt_dba
extraVolumeMounts:
- mountPath: /vt/backups
name: vitess-backups
extraVolumes:
- name: vitess-backups
persistentVolumeClaim:
claimName: ope-vitess-backup
replicas: 2
resources:
limits:
memory: 128Mi
requests:
cpu: 200m
memory: 128Mi