spec:
size: 3
This will use default version that etcd operator chooses.
spec:
size: 3
version: "3.1.8"
spec:
size: 3
pod:
nodeSelector:
diskType: ssd
antiAffinity: true
spec:
size: 3
pod:
resources:
limits:
cpu: 300m
memory: 200Mi
requests:
cpu: 200m
memory: 100Mi
See example .
spec:
size: 3
backup:
backupIntervalInSecond: 1800
maxBackups: 5
storageType: "S3"
s3:
s3Bucket: <S3-bucket-name>
awsSecret: <aws-secret-name>
prefix: <S3-prefix>
See backup config for how to set up S3 related configurations.
If a cluster cluster-a
was created with backup, but deleted or failed later on,
we can recover the cluster as long as the PV still exists.
Note that delete cluster-a
Cluster resource first if it still exists.
Here's an example:
metadata:
name: "cluster-a"
spec:
size: 3
backup:
backupIntervalInSecond: 300
maxBackups: 5
storageType: "PersistentVolume"
pv:
volumeSizeInMB: 512
restore:
backupClusterName: "cluster-a"
storageType: "PersistentVolume"
Same as above but using "S3" as backup storage. The previous s3 backup must exist in order for the restore to work; otherwise, the new cluster doesn't boot.
metadata:
name: "cluster-a"
spec:
size: 3
backup:
backupIntervalInSecond: 300
maxBackups: 5
storageType: "S3"
s3:
s3Bucket: <S3-bucket-name>
awsSecret: <aws-secret-name>
prefix: <S3-prefix>
restore:
backupClusterName: "cluster-a"
storageType: "S3"
etcd-operator can create a new cluster cluster-b
from the backup of another cluster cluster-a
. If backup exists, make sure the backup policy of cluster-b
has the same S3 configuration as those of cluster-a
; etcd operator would also use that config to access snapshots of cluster-a
.
metadata:
name: "cluster-a"
spec:
size: 3
backup:
backupIntervalInSecond: 300
maxBackups: 5
storageType: "S3"
s3:
s3Bucket: <S3-bucket-name>
awsSecret: <aws-secret-name>
prefix: <S3-prefix>
metadata:
name: "cluster-b"
spec:
size: 3
backup:
backupIntervalInSecond: 300
maxBackups: 5
storageType: "S3"
s3:
s3Bucket: <S3-bucket-name>
awsSecret: <aws-secret-name>
prefix: <S3-prefix>
restore:
backupClusterName: "cluster-a"
storageType: "S3"
If user wants to clone a new cluster cluster-b
from an existing cluster cluster-a
,
as long as backup exists, use the following example spec:
metadata:
name: "cluster-b"
spec:
size: 3
backup:
backupIntervalInSecond: 300
maxBackups: 5
storageType: "PersistentVolume"
pv:
volumeSizeInMB: 512
restore:
backupClusterName: "cluster-a"
storageType: "PersistentVolume"
spec:
size: 3
backup:
backupIntervalInSecond: 1800
maxBackups: 5
storageType: "ABS"
abs:
absContainer: <abs-container-name>
absSecret: <abs-secret-name>
See cluster TLS docs.