How to configure S3 backup storage using values.yaml file or helm chart options?
backup:
storages:
s3-storage:
type: s3
bucket: "my-backup"
region: default
endpointUrl: "https://s3.mystorage.local"
schedule:
- name: "daily-s3"
schedule: "0 3 * * *"
keep: 5
storage: s3-storage
type: full
Get an error:
How to fix this?
Hello everyone, I managed to set up that part in values, I even set up s3 parameters in values of pg operator, but to no avail.
storages:
- name: repo1
s3:
type: "s3"
bucket: "percona-backup"
.....
A backup pod is always running this:
msg="command to execute is [pgbackrest backup --type=full --db-host=10.42.0.184 --db-path=/pgdata/pg-db]"
I managed to run arguments using pgtask:
apiVersion: pg.percona.com/v1
kind: Pgtask
metadata:
labels:
pg-cluster: pg-db
pgouser: admin
name: pg-db-backrest-full-backup
spec:
name: pg-db-backrest-full-backup
parameters:
backrest-command: backup
# backup type can be:
# Differential: Create a backup of everything since the last full backup was taken
# --type=diff
# Full: Back up the entire database
# --type=full
# Incremental: Back up everything since the last backup was taken, whether it was full, differential, or incremental
# --type=incr
# backup retention can be:
# --repo1-retention-full=2 how many full backups to retain
# --repo1-retention-diff=4 how many differential backups to retain
# --repo1-retention-archive=6 how many sets of WAL archives to retain alongside
# the full and differential backups that are retained
backrest-opts: --type=full --repo1-retention-full=5 --repo1-type=s3 --repo1-s3-bucket=percona-backup --repo1-s3-endpoint=.......
backrest-s3-verify-tls: "false"
backrest-storage-type: "s3"
job-name: pg-db-backrest-full-backup
pg-cluster: pg-db
tasktype: backrest
but in this case, it’s not allowing me to use repo-s3-key in cmd line
I suppose it should be using environment variables from pg-db-backrest-repo-config (which I configured too) but it just isn’t. I don’t know what else to try. If anyone can help me it would be great. Thanks
Hi
Could you please share the helm CLI call on that?
Hello, helm command is the ususal
helm upgrade --install pg-db percona/pg-db --version 1.2.0 -n percona --create-namespace -f values-pg-db.yaml
And the values.yaml is:
backup:
image: perconalab/percona-postgresql-operator:main-ppg14-pgbackrest
# imagePullPolicy: Always
backrestRepoImage: perconalab/percona-postgresql-operator:main-ppg14-pgbackrest-repo
resources:
requests:
cpu: "200m"
memory: "48Mi"
# limits:
# cpu: "1"
# memory: "64Mi"
# affinity:
# antiAffinityType: preferred
volumeSpec:
size: 1G
accessmode: ReadWriteOnce
storagetype: dynamic
storageclass: "local-path"
# matchLabels: ""
storages:
- name: repo1
s3:
type: "s3"
bucket: "percona-backup"
region: "east-us"
uriStyle: "host"
verifyTLS: "false"
endpointUrl: "objectstorage.XXX.host"
schedule:
- name: "5min-backup"
schedule: "*/5 * * * *"
keep: 3
type: full
storage: local,s3
And I used this file to add pgbackrest params, to try and change the command in the backup pod
https://github.com/percona/percona-postgresql-operator/blob/main/deploy/backup/backup.yaml
I followed and tried everything from this doc:
It creates a local backup, and everything else is working fine, but I can’t get s3 upload to work.
Thank you for the detailed output. We appreciate it.
Helm chart for database cluster object has one small difference inside storage configuration - it’s not an array in terms of yaml but map.percona-helm-charts/README.md at pg-operator-1.2.0 · percona/percona-helm-charts · GitHub up. It has a small hint how to set backups storage as operator expects.
Hello, and thank you Ivan for your response.
helm upgrade --install pg-db percona/pg-db --version 1.2.0 -n percona --create-namespace -f values-pg-db.yaml \
--set backup.storages.my-s3.bucket=percona-backup \
--set backup.storages.my-s3.type=s3 \
--set backup.storages.my-s3.name=my-s3 \
--set backup.storages.my-s3.endpointUrl='objectstorage.XXXXXXXX.host' \
--set backup.storages.my-s3.region='west-us' \
--set backup.storages.storage0.verifyTLS=false \
--set backup.storages.storage0.uriStyle='path' \
--set bucket.key=IK6Oa8DM9HUgP3uC \
--set bucket.secret=Ec8xH9WU6CCDpXtMC2CNy1ko2b9hQJU9 \
--set backup.storages.my-local.type=local \
--set "backup.schedule[0].name=sat-night-backup" \
--set "backup.schedule[0].schedule=*/5 * * * *" \
--set "backup.schedule[0].keep=5" \
--set "backup.schedule[0].type=full" \
--set "backup.schedule[0].storage=s3" \
--set pgPrimary.volumeSpec.storageclass=local-path \
--set backup.volumeSpec.storageclass=local-path \
--set replicas.volumeSpec.storageclass=local-path
So now it does try to use S3, but stanza won’t create. Looks like it is joining the bucket name to minio host domain. I think I set it up correctly. So not sure what is going on
time="2022-08-05T08:18:51Z" level=info msg="stderr=[ERROR: [039]: HTTP request failed with 404 (Not Found):\n *** Path/Query ***:\n GET /?delimiter=%2F&list-type=2&prefix=backrestrepo%2Fpg-db-backrest-shared-repo%2Farchive%2Fdb%2F\n *** Request Headers ***:\n authorization: <redacted>\n content-length: 0\n host: percona-backup.objectstorage.XXXXX.host\n
Slavisa_Milojkovic:
uriStyle
HI
Try to use uriStyle: host instead. It should not append bucket name to the overall S3 URI.
Thank you very much Ivan, I made it work. The problem was that I used the wrong storage name in the code below.
Hi,
I am having a similar issue with setting up s3 backup on minio with values file.
finalizers:
crVersion: 2.3.1
repository: percona/percona-postgresql-operator
image: ""
imagePullPolicy: Always
postgresVersion: 15
pause: false
unmanaged: false
standby:
enabled: false
customTLSSecret:
name: ""
customReplicationTLSSecret:
name: ""
instances:
- name: service-1
replicas: 3
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 1
podAffinityTerm:
labelSelector:
matchLabels:
postgres-operator.crunchydata.com/data: postgres
topologyKey: kubernetes.io/hostname
walVolumeClaimSpec:
storageClassName: "default"
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 2Gi
dataVolumeClaimSpec:
storageClassName: "default"
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 2Gi
proxy:
pgBouncer:
replicas: 3
image: ""
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 1
podAffinityTerm:
labelSelector:
matchLabels:
postgres-operator.crunchydata.com/role: pgbouncer
topologyKey: kubernetes.io/hostname
backups:
pgbackrest:
image: ""
global:
repo1-retention-full: "14"
repo1-retention-full-type: time
repo2-retention-full: "14"
repo2-retention-full-type: time
repo2-s3-uri-style: path
repo2-s3-verify-tls: false
repo2-s3-key: "..."
repo2-s3-key-secret: "..."
repo2-retention-full: "14"
repo2-retention-full-type: time
repo2-path: /repo2
repoHost:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 1
podAffinityTerm:
labelSelector:
matchLabels:
postgres-operator.crunchydata.com/data: pgbackrest
topologyKey: kubernetes.io/hostname
priorityClassName: high-priority
topologySpreadConstraints:
- maxSkew: 1
topologyKey: my-node-label
whenUnsatisfiable: ScheduleAnyway
labelSelector:
matchLabels:
postgres-operator.crunchydata.com/pgbackrest: ""
manual:
repoName: repo1
options:
- --type=full
repos:
- name: repo1
schedules:
full: "* * * 7 *"
incremental: "* * * 5 *"
volume:
volumeClaimSpec:
storageClassName: "default"
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 1Gi
- name: repo2
s3:
bucket: "..."
endpoint: "..."
region: "default"
schedules:
full: "*/5 * * * *"
pmm:
enabled: false
image:
repository: percona/pmm-client
tag: 2.41.0
secret: cluster1-pmm-secret
serverHost: monitoring-service
secrets:
primaryuser: test123
postgres: test123
pgbouncer:
pguser: test123
I have been researching for a week and couldn’t make it work
Hello @mboncalo .
I do not see any secrets. Have you created any for your S3?
