pxc-db

pxc-db: A chart for installing Percona XtraDB Operator managed Databases

This chart implements Percona XtraDB Cluster deployment in Kubernets via Custom Resource object. The project itself can be found here:

• https://github.com/percona/percona-xtradb-cluster-operator

Pre-requisites

• PXC operator running in you K8S cluster
• Kubernetes 1.16+
• PV support on the underlying infrastructure - only if you are provisioning persistent volume(s).
• Helm v3

Custom Resource Details

• https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/

StatefulSet Details

• https://kubernetes.io/docs/concepts/workloads/controllers/statefulset/

Chart Details

This chart will:

• deploy a PXC database Pods (Custom Resource -> StatefulSet) for the further XtraDB Cluster creation in K8S.

Installing the Chart

To install the chart with the pxc release name using a dedicated namespace (recommended):

helm repo add percona https://percona.github.io/percona-helm-charts/
helm install my-db percona/pxc-db --version 0.1.17 --namespace my-namespace

The chart can be customized using the following configurable parameters:

Parameter	Description	Default
pause	Stop PXC Database safely	false
allowUnsafeConfigurations	Allows forbidden configurations like even number of PXC cluster pods	false
updateStrategy	Regulates the way how PXC Cluster Pods will be updated after setting a new image	SmartUpdate
upgradeOptions.versionServiceEndpoint	Endpoint for actual PXC Versions provider	https://check.percona.com/versions
upgradeOptions.apply	PXC image to apply from version service - recommended, latest, actual version like 8.0.19-10.1	8.0-recommended
upgradeOptions.schedule	Cron formatted time to execute the update	"0 4 * * *"
finalizers:delete-pxc-pods-in-order	Set this if you want to delete PXC pods in order on cluster deletion
finalizers:delete-proxysql-pvc	Set this if you want to delete proxysql persistent volumes on cluster deletion
finalizers:delete-pxc-pvc	Set this if you want to delete database persistent volumes on cluster deletion
pxc.size	PXC Cluster target member (pod) quantity. Can't even if allowUnsafeConfigurations is true	3
pxc.image.repository	PXC Container image repository	percona/percona-xtradb-cluster
pxc.image.tag	PXC Container image tag	8.0.22-13.1
pxc.autoRecovery	Enable full cluster crash auto recovery	true
pxc.imagePullSecrets	PXC Container pull secret	[]
pxc.annotations	PXC Pod user-defined annotations	{}
pxc.priorityClassName	PXC Pod priority Class defined by user
pxc.labels	PXC Pod user-defined labels	{}
pxc.readinessDelaySec	PXC Pod delay for readiness probe in seconds	15
pxc.livenessDelaySec	PXC Pod delay for liveness probe in seconds	300
pxc.forceUnsafeBootstrap	Order PXC Pods to override the previous Pod crash	false
pxc.configuration	User defined MySQL options according to MySQL configuration file syntax	``
pxc.resources.requests	PXC Pods resource requests	{"memory": "1G", "cpu": "600m"}
pxc.resources.limits	PXC Pods resource limits	{}
pxc.nodeSelector	PXC Pods key-value pairs setting for K8S node assingment	{}
pxc.affinity.antiAffinityTopologyKey	PXC Pods simple scheduling restriction on/off for host, zone, region	"kubernetes.io/hostname"
pxc.affinity.advanced	PXC Pods advanced scheduling restriction with match expression engine	{}
pxc.tolerations	List of node taints to tolerate for PXC Pods	[]
pxc.gracePeriod	Allowed time for graceful shutdown	600
pxc.podDisruptionBudget.maxUnavailable	Instruct Kubernetes about the failed pods allowed quantity	1
pxc.persistence.enabled	Requests a persistent storage (hostPath or storageClass) from K8S for PXC Pods datadir	true
pxc.persistence.hostPath	Sets datadir path on K8S node for all PXC Pods. Available only when pxc.persistence.enabled: true
pxc.persistence.storageClass	Sets K8S storageClass name for all PXC Pods PVC. Available only when pxc.persistence.enabled: true	-
pxc.persistence.accessMode	Sets K8S persistent storage access policy for all PXC Pods	ReadWriteOnce
pxc.persistence.size	Sets K8S persistent storage size for all PXC Pods	8Gi
pxc.disableTLS	Disable PXC Pod communication with TLS	false
pxc.certManager	Enable this option if you want the operator to request certificates from cert-manager	false
haproxy.enabled	Use HAProxy as TCP proxy for PXC cluster	true
haproxy.size	HAProxy target pod quantity. Can't even if allowUnsafeConfigurations is true	3
haproxy.image.repository	HAProxy Container image repository	percona/percona-xtradb-cluster-operator
haproxy.image.tag	HAProxy Container image tag	1.8.0-haproxy
haproxy.imagePullSecrets	HAProxy Container pull secret	[]
haproxy.annotations	HAProxy Pod user-defined annotations	{}
haproxy.priorityClassName	HAProxy Pod priority Class defined by user
haproxy.externalTrafficPolicy	Desire service to route external traffic to node-local or cluster-wide endpoints
haproxy.loadBalancerSourceRanges	Limit which client IP's can access the Network Load Balancer	[]
haproxy.serviceType	Specify what kind of Service you want	ClusterIP
haproxy.serviceAnnotations	Specify service annotations	{}
haproxy.labels	HAProxy Pod user-defined labels	{}
haproxy.readinessDelaySec	HAProxy Pod delay for readiness probe in seconds	15
haproxy.livenessDelaySec	HAProxy Pod delay for liveness probe in seconds	300
haproxy.forceUnsafeBootstrap	Order HAProxy Pods to override the previous Pod crash	false
haproxy.resources.requests	HAProxy Pods resource requests	{"memory": "1G", "cpu": "600m"}
haproxy.resources.limits	HAProxy Pods resource limits	{}
haproxy.nodeSelector	HAProxy Pods key-value pairs setting for K8S node assingment	{}
haproxy.affinity.antiAffinityTopologyKey	HAProxy Pods simple scheduling restriction on/off for host, zone, region	"kubernetes.io/hostname"
haproxy.affinity.advanced	HAProxy Pods advanced scheduling restriction with match expression engine	{}
haproxy.tolerations	List of node taints to tolerate for HAProxy Pods	[]
haproxy.gracePeriod	Allowed time for graceful shutdown	600
haproxy.podDisruptionBudget.maxUnavailable	Instruct Kubernetes about the failed pods allowed quantity	1
proxysql.enabled	Use ProxySQL as TCP proxy for PXC cluster	false
proxysql.size	ProxySQL target pod quantity. Can't even if allowUnsafeConfigurations is true	3
proxysql.image.repository	ProxySQL Container image repository	percona/percona-xtradb-cluster-operator
proxysql.image.tag	ProxySQL Container image tag	1.8.0-proxysql
proxysql.imagePullSecrets	ProxySQL Container pull secret	[]
proxysql.annotations	ProxySQL Pod user-defined annotations	{}
proxysql.priorityClassName	ProxySQL Pod priority Class defined by user
proxysql.externalTrafficPolicy	Desire service to route external traffic to node-local or cluster-wide endpoints
proxysql.loadBalancerSourceRanges	Limit which client IP's can access the Network Load Balancer	[]
proxysql.serviceType	Specify what kind of Service you want	ClusterIP
proxysql.serviceAnnotations	Specify service annotations	{}
proxysql.labels	ProxySQL Pod user-defined labels	{}
proxysql.readinessDelaySec	ProxySQL Pod delay for readiness probe in seconds	15
proxysql.livenessDelaySec	ProxySQL Pod delay for liveness probe in seconds	300
proxysql.forceUnsafeBootstrap	Order ProxySQL Pods to override the previous Pod crash	false
proxysql.resources.requests	ProxySQL Pods resource requests	{"memory": "1G", "cpu": "600m"}
proxysql.resources.limits	ProxySQL Pods resource limits	{}
proxysql.nodeSelector	ProxySQL Pods key-value pairs setting for K8S node assingment	{}
proxysql.affinity.antiAffinityTopologyKey	ProxySQL Pods simple scheduling restriction on/off for host, zone, region	"kubernetes.io/hostname"
proxysql.affinity.advanced	ProxySQL Pods advanced scheduling restriction with match expression engine	{}
proxysql.tolerations	List of node taints to tolerate for ProxySQL Pods	[]
proxysql.gracePeriod	Allowed time for graceful shutdown	600
proxysql.podDisruptionBudget.maxUnavailable	Instruct Kubernetes about the failed pods allowed quantity	1
proxysql.persistence.enabled	Requests a persistent storage (hostPath or storageClass) from K8S for ProxySQL Pods	true
proxysql.persistence.hostPath	Sets datadir path on K8S node for all ProxySQL Pods. Available only when proxysql.persistence.enabled: true
proxysql.persistence.storageClass	Sets K8S storageClass name for all ProxySQL Pods PVC. Available only when proxysql.persistence.enabled: true	-
proxysql.persistence.accessMode	Sets K8S persistent storage access policy for all ProxySQL Pods	ReadWriteOnce
proxysql.persistence.size	Sets K8S persistent storage size for all ProxySQL Pods	8Gi
logcollector.enabled	Enable log collector container	true
logcollector.image.repository	Log collector image repository	percona/percona-xtradb-cluster-operator
logcollector.image.tag	Log collector image tag	1.8.0-logcollector
pmm.enabled	Enable integration with Percona Monitoting and Management software	false
pmm.image.repository	PMM Container image repository	percona/pmm-client
pmm.image.tag	PMM Container image tag	2.12.0
pmm.serverHost	PMM server related K8S service hostname	monitoring-service
pmm.serverUser	Username for accessing PXC database internals	admin
backup.enabled	Enables backups for PXC cluster	true
backup.image.repository	Backup Container image repository	percona/percona-xtradb-cluster-operator
backup.image.tag	Backup Container image tag	1.8.0-pxc8.0-backup
backup.imagePullSecrets	Backup Container pull secret	[]
backup.pitr.enabled	Enable point in time recovery	false
backup.pitr.storageName	Storage name for PITR	s3-us-west-binlogs
backup.pitr.timeBetweenUploads	Time between uploads for PITR	60
backup.storages.fs-pvc	Backups storage configuration, where storages: is a high-level key for the underlying structure. fs-pvc is a user-defined storage name.
backup.storages.fs-pvc.type	Backup storage type	filysystem
backup.storages.fs-pvc.volume.persistentVolumeClaim.accessModes	Backup PVC access policy	["ReadWriteOnce"]
backup.storages.fs-pvc.volume.persistentVolumeClaim.resources	Backup Pod resources specification	{}
backup.storages.fs-pvc.volume.persistentVolumeClaim.resources.requests.storage	Backup Pod datadir backups size	6Gi
backup.schedule	Backup execution timetable	[]
backup.schedule.0.name	Backup execution timetable name	daily-backup
backup.schedule.0.schedule	Backup execution timetable cron timing	0 0 * * *
backup.schedule.0.keep	Backup items to keep	0 0 * * *
backup.schedule.0.storageName	Backup target storage	fs-pvc
secrets.passwords.root	Default user secret	insecure-root-password
secrets.passwords.xtrabackup	Default user secret	insecure-xtrabackup-password
secrets.passwords.monitor	Default user secret	insecure-monitor-password
secrets.passwords.clustercheck	Default user secret	insecure-clustercheck-password
secrets.passwords.proxyadmin	Default user secret	insecure-proxyadmin-password
secrets.passwords.pmmserver	Default user secret	insecure-pmmserver-password
secrets.passwords.operator	Default user secret	insecure-operator-password
secrets.tls	Not needed in case if you're using cert-manager. Structure expects keys ca.crt, tls.crt, tls.key and files contents encoded in base64.	{}

Specify parameters using --set key=value[,key=value] argument to helm install

Examples

Deploy a Cluster without a MySQL Proxy, no backups, no persistent disks

This is great for a dev cluster as it doesn't require a persistent disk and doesn't bother with a proxy, backups, or TLS.

$ helm install dev  --namespace pxc . \
    --set proxysql.enabled=false --set pxc.disableTLS=true \
    --set pxc.persistence.enabled=false --set backup-enabled=false

Deploy a cluster with certificates provided by Cert Manager

First you need a working cert-manager installed with appropriate Issuers set up. Check out the JetStack Helm Chart to do that.

By setting pxc.certManager=true we're signaling the Helm chart to not create secrets,which will in turn let the operator know to request appropriate certificate resources to be filled by cert-manager.

$ helm install dev  --namespace pxc . --set pxc.certManager=true

Deploy a production grade cluster

The pxc-database chart contains an example production values file that should set you well on your path to running a production database. It is not fully production grade as there are some requirements for you to provide your own secrets for passwords and TLS to be truly production ready, but it does provide comments on how to do those parts.

$ helm install prod --file production-values.yaml --namespace pxc .