Release v1.7.0

Release v1.7.0

• Enhancement & New Features
  • Support restoring from Alibaba Cloud PolarDB-X backup set
  • Support backup set compress
  • Support Huawei Cloud OBS as storage provider by s3 protocol
  • Change the failure threshold to 300s when startuping cn engine
  • Add spec info to backup set of standard instance
  • Skip validation when syncing spec from backup set
  • Support the latest version of DN
• Bug Fix
  • Fix polardbx_root account type
  • Fix delete expired files wrongly by s3 protocol

Release v1.6.2

Release v1.6.2

• Enhancement & New Features
  • Change the default image repo to polardbx-opensource-registry.cn-beijing.cr.aliyuncs.com/polardbx
• Bug Fix
  • Fix rebuild failure when the polardbx cluster is deployed with data-log-separation mode
  • Fix the config LocalExpireLogSeconds does not work as expected for binlog backup
  • Fix hpfs component fails to start when cpuset binding is not supported in the deployment environment

Release v1.6.1

Release v1.6.1

• Enhancement & New Features

  • Support download controller-gen when make manifest in the Makefile by @free6om in #40
  • Support tolerations #4 by @free6om in #40
  • Change the grace period to 300 seconds when deleting DN pod

• Bugfix

  • Fix cross Kubernetes cluster restore failure
  • Fix CN & CDC restart when only changing DN image

New Contributors

• @free6om made their first contribution in #39

Suggestions

• Before upgrading DN to 8.0.32, please upgrade the operator version to v1.6.1, and then make sure that .spec.config.cn.static.RPCProtocolVersion of PolarDXCluster is 2

Full Changelog: v1.6.0...v1.6.1

Release v1.6.0

Release v1.6.0

• Enhancement & New Features
  • Support Columnar Node for HTAP
  • Support CPU isolation
  • Support clinic component for efficient diagnosis
  • Compatible DN with MySQL 8.0.32
  • Support backup and restore for PolarDB-X Standard Edition
  • Support TDE (Transparent Data Encryption)
  • Improve the stability of backup and restore
  • Support cleanup of remote backup file when deleting backup object
  • Upgrade default RPCProtocolVersion to 2.0
  • Support collecting slow_detail.log instead of slow.log for CN slow log
  • Remove htc monitor from CN exporter
  • Enhance argument parsing and cleanup of the backup object during restoration by backupSetPath
• Bug Fix
  • Fix duplicated DN in GSM table: storage_info
  • Fix readonly PolarDB-X Cluster scale out failure
  • Fix extra xstore creation in the readonly PolarDB-X Cluster when DN replica count is one
  • Fix show ds issue when scale in
  • Fix wrong filestream filepath when the host filepath is not /filestream
  • Fix timezone of Operator differs from local
  • Fix update invalid when PolarDBXBackupSchedule is created with misformatted crontab
  • Fix serviceType of restored PolarDB-X Cluster differs from original Cluster when syncSpecWithOriginalCluster is true
  • Fix follower replication terminate when backup process exits abnormally
  • Fix backup failure when readonly PolarDB-X Cluster exists
  • Fix backup failure caused by waiting SeekCpJob process
  • Fix binlog purge disabled when backup fails
  • Fix fileno leak when backuping binlogs
  • Fix restore failure when all the DN pods are deleted
  • Fix status loss of backup related objects caused by misuisng Kubernetes client
  • Fix the issue of backup failures to MinIO buckets with TLS encryption

PolarDBX-Operator Release v1.6.0-beta

Release v1.6.0-beta

• Enhancement & New Features
  • Support Columnar Node for HTAP
  • Support CPU isolation
  • Support clinic component for efficient diagnosis
  • Compatible DN with MySQL 8.0.32
  • Support backup and restore for PolarDB-X Standard Edition
  • Support TDE (Transparent Data Encryption)
  • Improve the stability of backup and restore
  • Support cleanup of remote backup file when deleting backup object
  • Upgrade default RPCProtocolVersion to 2.0
  • Support collecting slow_detail.log instead of slow.log for CN slow log
  • Remove htc monitor from CN exporter
  • Enhance argument parsing and cleanup of the backup object during restoration by backupSetPath
• Bug Fix
  • Fix duplicated DN in GSM table: storage_info
  • Fix readonly PolarDB-X Cluster scale out failure
  • Fix extra xstore creation in the readonly PolarDB-X Cluster when DN replica count is one
  • Fix show ds issue when scale in
  • Fix wrong filestream filepath when the host filepath is not /filestream
  • Fix timezone of Operator differs from local
  • Fix update invalid when PolarDBXBackupSchedule is created with misformatted crontab
  • Fix serviceType of restored PolarDB-X Cluster differs from original Cluster when syncSpecWithOriginalCluster is true
  • Fix follower replication terminate when backup process exits abnormally
  • Fix backup failure when readonly PolarDB-X Cluster exists
  • Fix backup failure caused by waiting SeekCpJob process
  • Fix binlog purge disabled when backup fails
  • Fix fileno leak when backuping binlogs
  • Fix restore failure when all the DN pods are deleted
  • Fix status loss of backup related objects caused by misuisng Kubernetes client
  • Fix the issue of backup failures to MinIO buckets with TLS encryption

PolarDBX-Operator Release v1.5.0

Release v1.5.0

• Enhancement & New Features
  • Support PolarDB-X Standard(XStore) dashboard
  • Remove Cluster IP dependency on the paxos communication improving some performance
  • Support cross-cluster point-in-time recovery(PITR)
  • Support collecting DN logs including sql, error, slow logs
  • Support uploading and downloading backupset by Simple Storage Service (S3)
  • Support creating cdc groups with different cluster types
• Bug Fix
  • Fix failing to delete the cluster when its gms is not available
  • Fix failing to delele rebuild job when the fromPod does not exist
  • Fix disk util alert issue. #21
  • Fix helm template issue when set tolerations. #25
  • Fix failing to create backup when serviceName of PolarDBXCluster differs from its name

PolarDBX-Operator Release v1.4.1

• Enhancement & New Features
  • Support PolarDB-X Engine 2.0.0
  • Remove single download proxy for consensus logs in point-in-time recovery(PITR)
• Bug Fix
  • Fix readonly pxc when shareGms is true
  • Fix Compute Nodes metadata inconsistency when upgrading frequently

PolarDBX-Operator Release v1.4.0

• Enhancement & New Features
  • Support rebuild dn pod automatically
  • Support consensus log backup
  • Support point-in-time recovery(PITR)
  • Support backup schedule
  • Support restore without backup object
  • Support debug runmode of PolarDB-X cluster
  • Support default PrometheusRule for polardbx cluster.
  • Use Cluster IP as the default communication address among dn pods
  • Use container network as the default network type
• Bug Fix
  • Fix backup failure on follower when using non-host network
  • Fix grafana cpu and mem usage issue for kube-state-metrics 2.3.0
  • Fix restore issue when binlog checksum is none.

Release v1.4.0-beta

• Enhancement & New Features
  • Support consensus log backup
  • Support point-in-time recovery(PITR)
  • Support backup schedule
  • Support restore without backup object
  • Support rebuild dn follower automatically when Slave_SQL_Running becomes No due to Last_SQL_Error
  • Support debug runmode of PolarDB-X cluster
  • Support default PrometheusRule for polardbx cluster
  • Use Cluster IP as the default communication address among dn pods
  • Use container network as the default network type
• Bug Fix
  • Fix backup failure on follower when using non-host network
  • Fix grafana cpu and mem usage issue for kube-state-metrics 2.3.0
  • Fix restore issue when binlog checksum is none
  • Fix cdc crash during backup caused by sending heartbeat using XA transaction

PolarDB-X Operator Release v1.3.0

• Enhancement & New Features
  • Support backup and restore.
  • Support multiple architecture: linux/amd64, linux/arm64.
  • Support log(audit, error) collection.
  • Support read only cluster.
  • Support parameter template and configuration.
  • Support XStore node rebuild.
  • Support splitting dn data and log into different directories.
  • Support archiving cold data into OSS and NFS.
• Bug Fix
  • Fix mycnfOverwrite not working issue.
  • Fix issue of cn topology rules not working in host network mode.
  • Fix XStore node not ready when ip changed in container network mode.
  • Fix logstash oom
  • Fix rebuild job does continue, when the instance is upgrading
  • Fix failing to create readonly pxc when dnReplicas is less than 2
  • Fix xstore topology of restored instance not in sync with original instance
  • Fix binlog purge lock not working during backup
  • Fix change leader command issue for xsl.
  • Fix backup progress hang in binlogbackuping phase.