tencent cloud

PrevNext

Feedback

search by keyword

Recent Pages

Documentation
Tencent Kubernetes Engine
    Documentation
    Download PDF

    TKE Kubernetes Revision Version History

    Last updated: 2024-05-28 10:24:39
    Download PDF

      TKE kubernetes 1.28.3 revisions

      Time
      Version
      Update Contents
      2024-04-24
      v1.28.3-tke.3
      lb vip binding issue fixed. (kube-proxy)
      2024-04-10
      v1.28.3-tke.2
      Merge PR119561 to fix the issue where kubelet using gpu might cause panic. (kubelet)
      Fix issues with scheduling pods with eni-ip to super nodes. (kube-scheduler)
      2024-03-06
      v1.28.3-tke.1
      Support specified Pods during scale-down. (kube-controller-manager)
      Avoid scheduling Pods using cbs to external CHC Nodes. (kube-scheduler)
      Pods with added annotation support mounting cgroupfs. (kubelet)
      Support native node scaling feature. (kube-scheduler)
      Correctly display disk and network metrics when runtime is containerd. (kubelet)
      Super nodes support DaemonSet. (kube-apiserver/kube-controller-manager)
      Optimize super node HPA. (kube-controller-manager)
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      Extend the apiserver's self-signed certificate validity from 1 year to 5 years. (kube-apiserver)
      Super nodes allow scheduling Pods in TKE trunking ENI mode. (kube-scheduler)

      TKE kubernetes 1.26.1 revisions

      Time
      Version
      Update Contents
      2024-04-10
      v1.26.1-tke.6
      Merge PR119561 to fix the issue where kubelet using gpu might cause panic. (kubelet)
      Merge PR113145 to fix the issue where forced pod deletion might leave behind a sandbox, preventing pod reconstruction. (kubelet)
      2024-02-27
      v1.26.1-tke.5
      Super nodes allow scheduling Pods in TKE trunking ENI mode. (kube-scheduler)
      2024-02-01
      v1.26.1-tke.4
      Super node supports Reserved Specification Awareness. (kube-scheduler)
      Merge PR119605 to fix incomplete pod information in kubelet prober_probe_total metric. (kubelet)
      kube-proxy Optimization. (kube-proxy)
      job-controller supports settings for multi-worker parameters --concurrent-job-syncs. (kube-controller-manager)
      2023-12-04
      v1.26.1-tke.3
      Pods with added annotation support mounting cgroupfs. (kubelet)
      Merge PR119706 to fix the issue of Pods using device plugin resources being reconstructed when kubelet restarts. (kube-apiserver)
      When deleting resources, add userAgent reporting and record detailed logs. (kube-apiserver)
      2023-10-12
      v1.26.1-tke.2
      Support in-place updates for Pod resources. (kube-apiserver, kubelet)
      Merge PR117550 to add --concurrent-cron-job-syncs parameter to set the worker quantity of the cronjob controller. (kube-controller-manager)
      Error in fixing the scheduler iterator, avoiding score exceeding the range of a single plugin due to adding duplicate scorePlugin. (kube-scheduler)
      Read cache optimization. (kube-apiserver)
      Extend the self-signed certificate validity from 1 year to 5 years. (kube-apiserver)
      2023-03-30
      v1.26.1-tke.1
      Support specified Pods during scale-down. (kube-controller-manager)
      Avoid scheduling Pods using cbs to external CHC Nodes. (kube-scheduler)
      Fix issues with LB health checks and IPVS. (kube-proxy)
      Optimize super node HPA. (kube-controller-manager)
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      Support native node scaling feature. (kube-scheduler)
      Enhancements to kube-apiserver's priority and fairness features. (kube-apiserver)
      APF Enhancement: Sort by pl name upon dump_priority_levels, and add PL debug information, which is merged into community PR112393.
      Add metrics for list/watch. (kube-apiserver)
      Support exec/logs commands for super nodes and DaemonSet Pods. (kube-apiserver)
      Add metrics for network/disk, etc., for the Containerd runtime. (kubelet)
      When kube-controller-manager performs pod gc, prioritize deleting Evicted pods. (kube-controller-manager)

      TKE kubernetes 1.24.4 revisions

      Note:
      Kubernetes 1.24 has removed support for Docker through Dockershim. For the container runtime on new nodes, please use Containerd. Images built with Docker can continue to be used. For more information, please refer to Dockershim Removal FAQ.
      Time
      Version
      Update Contents
      2024-04-10
      v1.24.4-tke.14
      Merge PR119561 to fix the issue where kubelet using gpu might cause panic. (kubelet)
      Merge PR113145 to fix the issue where forced pod deletion might leave behind a sandbox, preventing pod reconstruction. (kubelet)
      2024-02-27
      v1.24.4-tke.13
      Sandbox optimization for super nodes retained. (kube-scheduler)
      Super nodes allow scheduling Pods in TKE trunking ENI mode. (kube-scheduler)
      Optimization of computing resource policies for super nodes. (kube-scheduler)
      2024-02-01
      v1.24.4-tke.12
      kube-proxy Optimization. (kube-proxy)
      job-controller support settings for multi-worker parameters --concurrent-job-syncs. (kube-controller-manager)
      2023-12-04
      v1.24.4-tke.11
      Pods with added annotation support mounting cgroupfs. (kubelet)
      When deleting resources, add userAgent reporting and record detailed logs. (kube-apiserver)
      2023-10-12
      v1.24.4-tke.10
      Port CVE-2022-3162. (kube-apiserver)
      Read cache optimization. (kube-apiserver)
      Extend the self-signed certificate validity from 1 year to 5 years. (kube-apiserver)
      Merge PR116327 to optimize memory usage in GetList. (kube-apiserver)
      Enhance monitoring metrics collection. (kube-scheduler)
      2023-07-21
      v1.24.4-tke.9
      Enable kube-proxy to detect changes in node PodCIDRs and ensure consistency through restarts upon changes. (kube-proxy)
      Fix the issue where kubelet may panic when assigning devices to pods. (kubelet)
      Error in fixing the scheduler iterator, avoiding score exceeding the range of a single plugin due to adding duplicate scorePlugin. (kube-scheduler)
      Port CVE-2022-3294 (kube-apiserver)
      2023-05-30
      v1.24.4-tke.7
      Support pod resource in-place updates. (kube-apiserver, kube-scheduler, kubelet)
      Optimize scheduling logs. (kube-scheduler)
      Merge PR117550, to add --concurrent-cron-job-syncs parameter to set the worker count for the cronjob controller. (kube-controller-manager)
      2023-03-17
      v1.24.4-tke.6
      Ignore TKEDefaultQuota conflict errors, and increase worker count to resolve resource quota evaluates timeout errors. (kube-apiserver)
      Support specified Pods during scale-down. (kube-controller-manager)
      Prevent Pods using CBS from being scheduled to external CHC nodes. (kube-scheduler)
      Merge PR110951 to avoid the issue where sporadic PV mount operations get stuck, causing Pods associated with that PV to remain in the Creating state. (kubelet)
      Merge PR111635 to fix the issue where ipvs real server weight is 0 after kube-proxy restarts. (kube-proxy)
      Support native node scaling feature. (kube-scheduler)
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      Enhancements to kube-apiserver priority and fairness features: (kube-apiserver) a. Support rate limiting based on userAgent, support Token Bucket for rate limiting, support rate limiting, circuit breaking for any type of component. b. When dumping priority levels, sort by pl name; add PL debug information: Total Requests Processed, Total Requests Rejected, Total Requests Timed Out, Total Requests Cancelled.
      Record metrics when list/watch is too old, too large, and Excessive Objects events (kube-apiserver): watch_too_old_objects_events_total , list_too_large_objects_events_total , list_too_many_objects_events_total , watch_too_many_objects_events_total
      Support network metrics for Containerd runtime. (kubelet)
      Support Tencent Cloud CBS CSI migration. (kube-apiserver,kube-controller-manager,kubelet)
      When kube-controller-manager performs pod gc, prioritize deleting Evicted pods. (kube-controller-manager)
      2023-02-16
      v1.24.4-tke.5
      Support exec/logs commands for super node DaemonSet Pod. (kube-apiserver)
      Super node supports fixed IP address. (kube-scheduler)
      Port 1.26 to support multiple HPAController workers. (kube-controller-manager)
      Port the community heuristic feature to set watch channel size and support setting certain resources' watch channel size via env. (kube-apiserver)
      Optimize self-adaptive node iptables mode Script, and fix misjudgments in some cases. (kube-proxy)
      2022-12-08
      v1.24.4-tke.3
      Support Disk Usage Metrics upon Containerd runtime. (kubelet)
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      Support TKE Serverless super nodes. (kube-controller-manager,kube-scheduler)
      2022-05-07
      v1.24.4-tke.1
      Update the Startup mode for running kube-proxy in an Image manner, and automatically adapt to the iptables running mode of the node, to support the operating system that runs iptables in nf_tables mode by default. (kube-proxy)
      Allows special IP ranges used by TKE managed clusters. (kube-apiserver)
      revert pr63066, fixing the issue with LB health check and IPVS. (kube-proxy)
      Merge PR112299 to optimize the gzip compression level in apiserver.

      TKE kubernetes 1.22.5 revisions

      Time
      Version
      Update Contents
      2024-04-10
      v1.22.5-tke.24
      Merge PR119561 to fix the issue where kubelet using gpu might cause panic. (kubelet)
      Merge PR106570 to fix the issue with kubelet setting cpushares. (kubelet)
      2024-02-27
      v1.22.5-tke.23
      Super nodes allow scheduling Pods in TKE trunking ENI mode. (kube-scheduler)
      Optimization of computing resource policies for super nodes. (kube-scheduler)
      2024-02-01
      v1.22.5-tke.22
      job-controller supports settings for multi-worker parameters --concurrent-job-syncs. (kube-controller-manager)
      2023-12-04
      v1.22.5-tke.21
      Port pr108753 to fix the issue where pod memory and other metrics couldn't be retrieved through kubelet /stats/summary when using systemd cgroup. (kubelet)
      Pods with added annotation support mounting cgroupfs. (kubelet)
      When deleting resources, add userAgent reporting and record detailed logs. (kube-apiserver)
      2023-10-12
      v1.22.5-tke.20
      Extend the self-signed certificate validity from 1 year to 5 years. (kube-apiserver)
      Merge PR116327 to optimize memory usage in GetList. (kube-apiserver)
      Merge PR107900 to fix the issue where static pods may fail to start under certain scenarios. (kubelet)
      Revert the fix for the upgrade issue caused by ubuntu16.04lxcfs, as the fix is no longer needed and may lead to e2e test failures. (kubelet)
      Port CVE-2022-3162. (kube-apiserver)
      2023-08-02
      v1.22.5-tke.19
      Read cache optimization. (kube-apiserver)
      2023-07-21
      v1.22.5-tke.18
      Fix the issue where kubelet may panic when assigning devices to pods. (kubelet)
      Fix the bug in the super node's definition policy. (kube-scheduler)
      Port CVE-2022-3294 (kube-apiserver)
      2023-05-26
      v1.22.5-tke.16
      Port PR105606&PR105872 to fix the goroutine leak issue caused by DeleteCollection. (kube-apiserver)
      Port PR117550, to add --concurrent-cron-job-syncs parameter to set the worker count for the cronjob controller. (kube-controller-manager)
      2023-04-20
      v1.22.5-tke.15
      Skip the creation of new container-related files when cadvisor collects metrics to avoid reporting null values and causing breakpoints. (kubelet)
      Merge PR112913 to fix the issue with cpu manager calculating allocatable cpusets. (kubelet)
      Improve the processing logic for in-place Pod configuration changes. (kube-apiserver,kube-scheduler,kubelet)
      Merge PR107670 to fix a potential timing bug upon Pod deletion that may lead to sandbox leakage. (kubelet)
      2023-03-24
      v1.22.5-tke.14
      Merge PR108202 to fix the missing nodeName field in EndpointSlice after upgrading from 1.20 to 1.22. (kube-apiserver)
      Merge PR110245 to add a fallback logic to kube-proxy that uses endpoint.DeprecatedTopology[v1.LabelHostname] when endpoint.NodeName is empty, to avoid issues of service inaccessibility after upgrading from 1.20 to 1.22. (kube-proxy)
      2023-03-17
      v1.22.5-tke.13
      Merge PR110951 to avoid the issue where sporadic PV mount operations get stuck, causing Pods associated with that PV to remain in the Creating state. (kubelet)
      Fix the issue where some GET requests were not logged as WATCH. (kube-apiserver)
      When kube-controller-manager performs pod gc, prioritize deleting Evicted pods. (kube-controller-manager)
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      2023-03-06
      v1.22.5-tke.12
      Merge PR111635 to fix the issue where ipvs real server weight is 0 after kube-proxy restarts. (kube-proxy)
      2023-02-16
      v1.22.5-tke.11
      Super node supports fixed IP address. (kube-scheduler)
      Port 1.26 to support multiple HPAController workers. (kube-controller-manager)
      Port the community heuristic feature to set watch channel size and support setting certain resources' watch channel size via env. (kube-apiserver)
      Port PR108366, and fix kubelet reporting pod status as finished too early, leading to insufficient resources when new pods are running. (kubelet)
      Fix inotify leaks causing kubelet memory leaks. (kubelet)
      Remove logs related to watch_too_old_objects_events_total, list_too_many_objects_events_total, watch_too_many_objects_events_total, avoid printing excessive information. (kube-apiserver)
      Optimize self-adaptive node script in iptables mode, and fix misjudgments in some cases. (kube-proxy)
      2023-01-12
      v1.22.5-tke.9
      Support native node scaling feature. (kube-scheduler)
      Add a delegated_to_storage Tag to the list_too_many_objects_events_total metric, indicating whether the request is forwarded to etcd. (kube-apiserver)
      2023-01-05
      v1.22.5-tke.8
      Enhancements to kube-apiserver priority and fairness features: (kube-apiserver)
      a. When applying rate limiting, consider the number of objects of the list request's resource type. The more objects, the lower the concurrent QPS, and the less resources APIServer will occupy, significantly reducing resource usage and effectively enhancing cluster stability to avoid avalanches. If business components issue a large number of list requests, upgrades may result in increased latency, with APIServer returning rate limiting and retries, etc.
      b. Support rate limiting based on userAgent, token bucket rate limiting, and rate limiting or circuit breaking for any type of component.
      c. Support rate limiting for update requests related to objects involved in a watch to avoid generating too many update events at once.
      Super nodes support running DaemonSet Pods when needed. (kube-controller-manager)
      Support exec/logs commands for super node DaemonSet Pod. (kube-apiserver)
      Record metrics when encountering list/watch too old, too large, or too many objects events (kube-apiserver): watch_too_old_objects_events_total, list_too_large_objects_events_total, list_too_many_objects_events_total, watch_too_many_objects_events_total
      Merge PR108754 to fix the issue of reporting not registered error when occasionally mounting ConfigMap/Secrets volumes. (kubelet)
      Support in-place updates for Pod resources. (kube-apiserver, kubelet)
      2022-12-08
      v1.22.5-tke.7
      Fix the problem of container network interface default name being set to eth1. (kubelet)
      2022-11-09
      v1.22.5-tke.6
      Optimize monthly subscription super node scheduling. (kube-scheduler)
      Optimize cloud-native node stability. (kubelet)
      Optimize local replica count scheduling for super nodes.
      Add QosIgnoreInitContainer switch, which ignores resources of init containers when setting pod QoS upon being turned on. (kube-apiserver, kubelet)
      2022-10-13
      v1.22.5-tke.5
      Serverless clusters support daemonset. (kube-apiserver)
      Merge PR112299 to optimize the gzip compression level in apiserver. (kube-apiserver)
      Fixed CVE-2022-3172. (kube-apiserver)
      Add calling_webhook_timeout_error metric. (kube-apiserver)
      Ignore TKEDefaultQuota conflict errors, increase worker count to resolve resource quota evaluates timeout errors. (kube-apiserver)
      Merge PR110294 to fix the issue where the Job activeDeadlineSeconds setting was ineffective. (kube-controller-manager)
      Merge PR111773 to fix memory leak issues during scheduler preemption. (kube-scheduler)
      Support containerd disk metrics. (kubelet)
      Merge PR108831 to fix kubelet panic when multiple Pods are created simultaneously. (kubelet)
      Support specified Pods during scale-down. (kube-controller-manager)
      Optimize daemonset pod scheduling performance by only processing assigned nodes. (kube-scheduler)
      Set CBS CSI Migration to completed status. (kube-controller-manager)
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      Optimize HPA for super nodes in Serverless Clusters. (kube-controller-manager)
      2022-05-07
      v1.22.5-tke.1
      Allows special IP ranges used by TKE managed clusters. (kube-apiserver)
      Restore pr63066 to fix the issue with LB health check and IPVS. (kube-proxy)
      Merge pr90260 to fix the missing Cluster Network Monitoring issue in containerd clusters. (kubelet)
      Fix the Exit Issues of Pods caused by the upgrade of lxcfs under ubuntu16. (kubelet)
      Prevent Pods using CBS from being scheduled to external CHC nodes. (kube-scheduler)
      Support Tencent Cloud CBS CSI Migration. (kube-controller-manager,kubelet)
      Merge pr106906 to detect whether the subpath of the Network Storage Volume has been deleted, to avoid Pods remaining in terminating Status. (kubelet)
      Update the Startup mode for running kube-proxy as an image, automatically adapting to the iptables Running Mode of the node, to support Operating Systems that run iptables in nf_tables mode by default. (kube-proxy)

      TKE kubernetes 1.20.6 revisions

      Time
      Version
      Update Contents
      2024-04-10
      v1.20.6-tke.43
      Customizable Eviction Delay by Definition, determining the order of availability zone eviction. (kube-scheduler)
      2024-02-27
      v1.20.6-tke.42
      Sandbox optimization for super nodes retained. (kube-scheduler)
      Workload Mounted Disk Performance Optimization. (kube-controller-manager)
      Super nodes allow scheduling Pods in TKE trunking ENI mode. (kube-scheduler)
      Super Node Computing Resource Policy Optimization. (kube-scheduler)
      2024-02-01
      v1.20.6-tke.41
      Native Node Support for CVM Mode. (kubelet)
      Merge PR102014 to fix the issue that CPUs bound to init containers cannot be used after release. (kubelet)
      job-controller support settings for multi-worker parameters --concurrent-job-syncs. (kube-controller-manager)
      2023-12-04
      v1.20.6-tke.40
      Pods with added annotation support mounting cgroupfs. (kubelet)
      scheduler patch pod optimization. (kube-scheduler)
      When deleting resources, add userAgent reporting and record detailed logs. (kube-apiserver)
      2023-10-12
      v1.20.6-tk.39
      Extend the self-signed certificate validity from 1 year to 5 years. (kube-apiserver)
      Revert the fix for the upgrade issue caused by ubuntu16.04lxcfs, as the fix is no longer needed and leads to e2e test failures. (kubelet)
      2023-08-02
      v1.20.6-tke.38
      Merge PR103320, Bump SMD to v4.1.2 to pick up #102749 fix.
      Read cache optimization. (kube-apiserver)
      2023-07-20
      v1.20.6-tke.37
      Merge PR116327, Optimize memory usage of List. (kube-apiserver)
      Make hpa controller's qps/burst configurable. (kube-controller-manager)
      Port PR107831, Include pod UID in kubelet's secret/configmap cache. (kubelet)
      Reinforce cadvisor collection. (kubelet)
      Optimize the performance of the attach detach controller and accelerate processing speed during high-frequency node updates. (kube-controller-manager)
      Merge CVE-2022-3294 (kube-apiserver)
      Fix the bug in the super node's Definition policy. (kube-scheduler)
      2023-05-26
      v1.20.6-tke.34
      Fix the issue where the watch_cache_capacity metrics for CRD instances are all Unstructured in their resource names, making them indistinguishable and affecting each other. (kube-apiserver)
      Port PR116167, prioritize deleting Evicted pods during pod GC. (kube-controller-manager)
      Port PR97845, add apiserver_terminated_watchers_total metric, representing the number of watchers closed due to not being processed in time. (kube-apiserver)
      Port PR105606&PR105872 to fix the goroutine leak issue caused by DeleteCollection. (kube-apiserver)
      Port PR117550, adding --concurrent-cron-job-syncs parameter to set the worker count for the cronjob controller. (kube-controller-manager)
      Improve the processing logic for in-place Pod configuration changes. (kube-apiserver,kube-scheduler,kubelet)
      2023-03-10
      v1.20.6-tke.33
      Merge PR110951, to avoid the issue where sporadic PV mount operations get stuck, causing Pods associated with that PV to remain in the Creating state. (kubelet)
      Merge PR106295, Device Manager reads the status file while adapting to both v1 and v2 formats. (kubelet)
      Fix the issue where some GET requests were not logged as WATCH. (kube-apiserver)
      2023-03-06
      v1.20.6-tke.32
      Merge PR111635 to fix the issue where ipvs real server weight is 0 after kube-proxy restarts. (kube-proxy)
      2023-02-13
      v1.20.6-tke.31
      Port 1.26 to support multiple HPAController workers. (kube-controller-manager)
      Port the community heuristic feature to set watch channel size and support setting certain resources' watch channel size via env. (kube-apiserver)
      Fix inotify leaks causing kubelet memory leaks. (kubelet)
      Remove logs related to watch_too_old_objects_events_total, list_too_many_objects_events_total, watch_too_many_objects_events_total, avoid printing excessive information. (kube-apiserver)
      Score super nodes based on the super node eks.tke.cloud.tencent.com/eklet-priority annotation. (kube-scheduler)
      Fix the bug with super node's fixed IP. (kube-scheduler)
      Optimize self-adaptive node iptables mode Script, fixing misjudgments in some cases. (kube-proxy)
      Fix the issue where the metric rest_client_rate_limiter_duration_seconds was unregistered. (kube-apiserver, kube-controller-manager, kube-scheduler, kubelet, kube-proxy)
      2023-01-05
      v1.20.6-tke.30
      Merge PR104833 to optimize the controller's lock in API Priority and Fairness to a read-write lock. (kube-apiserver)
      Super nodes support running DaemonSet Pods when needed. (kube-controller-manager)
      Merge PR102576, and no longer use RemoveAll when deleting legacy Pod volumes to avoid deleting volume-related data. (kubelet)
      Allow DaemonSet pods with specific annotations to be scheduled on super nodes. (kube-controller-manager)
      Add a switch for the local replica count policy for super nodes. (kube-scheduler)
      Enhancements to kube-apiserver priority and fairness features: (kube-apiserver)
      a. When applying rate limiting, consider the number of objects of the list request's resource type. The more objects, the lower the concurrent QPS, and the less resources APIServer will occupy, significantly reducing resource usage and effectively enhancing cluster stability to avoid avalanches. If business components issue a large number of list requests, upgrades may result in increased latency, with APIServer returning rate limiting and retries, etc.
      b. Support rate limiting based on userAgent, token bucket rate limiting, and rate limiting or circuit breaking for any type of component.
      c. Support rate limiting for update requests related to objects involved in a watch to avoid generating too many update events at once.
      Super node supports fixed IP address. (kube-scheduler)
      Add a switch for features related to super nodes. (kube-scheduler)
      Record metrics when list/watch is too old, too large, and excessive objects (kube-apiserver): watch_too_old_objects_events_total, list_too_large_objects_events_total, list_too_many_objects_events_total, watch_too_many_objects_events_total
      Add watch cache capacity metrics: watch_cache_capacity. (kube-apiserver)
      Use ConfigMap informer to optimize super node scheduling performance. (kube-scheduler)
      Support setting the minimum number of watch cache entries through the environment variable DEFAULT_LOWER_BOUND_CAPACITY. (kube-apiserver)
      Add AllowLocalConnection switch, which allows external requests to access local addresses when enabled. (kubelet)
      2022-11-09
      v1.20.6-tke.28
      Optimize monthly subscription super node scheduling. (kube-scheduler)
      Optimize cloud-native node stability. (kubelet)
      Fix native node in-place upgrade and downgrade kubelet abnormally exits. (kubelet)
      Optimize super node local replica number scheduling. (kube-scheduler)
      Add QosIgnoreInitContainer switch, which ignores resources of init containers when setting pod QoS upon being turned on. (kube-apiserver, kubelet)
      2022-10-13
      v1.20.6-tke.27
      Optimize scheduling policy for super node IP management. (kube-scheduler)
      Merge PR112299 to optimize the gzip compression level in apiserver. (kube-apiserver)
      Fix CVE-2022-3172. (kube-apiserver)
      Add calling_webhook_timeout_error metric. (kube-apiserver)
      Ignore TKEDefaultQuota conflict errors, and increase worker count to resolve resource quota evaluates timeout errors. (kube-apiserver)
      Optimize Serverless clusters local replica number scheduling. (kube-scheduler)
      Cancel fixed eip forced scheduling to the original super node. (kube-scheduler)
      Merge PR99324, and immediately reset after the kube-controller-manager token becomes invalid. (kube-controller-manager)
      Merge PR101155 to support setting multiple ServiceAccount Token Issuers. (kube-apiserver)
      Add StatusDetaching status for CBS Disks. (kube-controller-manager)
      Extend the timeout for updating resources when scheduling Static IP to super nodes, to prevent update failures during large-scale concurrency. (kube-scheduler)
      Serverless clusters support daemonset. (kube-apiserver)
      2022-09-07
      v1.20.6-tke.24
      Optimize scheduler preemption to avoid crashes. (kube-scheduler)
      Optimize super node scheduling. (kube-scheduler)
      Support in-place updates for Pod resources. (kube-apiserver, kubelet)
      Support Pods under kube-system to be scheduled to prepaid super nodes. (kube-scheduler)
      Optimize super node HPA. (kube-controller-manager)
      Merge PR110294 to fix the issue where the Job activeDeadlineSeconds setting was ineffective. (kube-controller-manager)
      Merge PR111773 to fix the memory leak issues during scheduler preemption. (kube-scheduler)
      Merge PR97348 to fix the issue where HPA scaling numbers are incorrect when StabilizationWindowSeconds is set. (kube-controller-manager)
      Optimize monthly subscription super node scheduling.
      Merge PR108831 to fix the issue where creating multiple Pods simultaneously causes kubelet to panic. (kubelet)
      Repair cronjob Pod creation failure when Pod name/uid is empty. (kube-controller-manager)
      2022-07-27
      v1.20.6-tke.21
      CBS supports native nodes. (kubelet)
      Optimize TKE Serverless super node HPA.
      2022-06-16
      v1.20.6-tke.20
      Support monthly subscription super nodes. (kube-scheduler, kube-controller-manager)
      When using docker and overlay2, disk usage is retrieved through fs quota to enhance performance. (kubelet)
      Optimize daemonset pod scheduling performance by only processing assigned nodes. (kube-scheduler)
      Optimize TKE Serverless scheduling. (kube-scheduler)
      TKE Serverless: Support mounting PVC after pod creation. (kube-scheduler)
      When scheduling to TKE Serverless nodes, you can ignore hugepages resources through a feature toggle. (kube-scheduler)
      2022-04-22
      v1.20.6-tke.17
      TKE Serverless: Support retaining sandbox features. (kube-scheduler)
      Merge pr101093 to avoid the issue where startupProbe does not run after container restart. (kubelet)
      2022-03-24
      v1.20.6-tke.16
      Resolve the issue where inline CSI and ephemeral generic temporary volumes become unavailable after upgrading to 1.20. (kube-apiserver, kube-controller-manager, kube-scheduler, kubelet, kubectl)
      2022-03-18
      v1.20.6-tke.15
      Support specifying a pod during scaling down. (kube-controller-manager)
      Merge pr106906 to check if the subpath of network storage volumes has been deleted, avoiding pods being stuck in terminating state. (kubelet)
      When scheduling anti-affinity based on hostname, ignore TKE Serverless super nodes. (kube-scheduler)
      Support upgrading from tke1.18 to 1.20. (kube-apiserver, kube-controller-manager, kubelet)
      Port pr108325 to fix the issue where kubelet might panic during startup due to the deletion of the pause container. (kubelet)
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      2022-01-20
      v1.20.6-tke.12
      TKE Serverless Rescheduling Optimization: For the same availability zone's super nodes that have been evicted, reduce their score. (kube-scheduler)
      The apiserver supports integration of ExternalName type external services. (kube-apiserver)
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      2021-12-09
      v1.20.6-tke.9
      Optimize TKE Serverless super node scheduling and HPA. (kube-controller-manager, kube-scheduler)
      Fix the inconsistency in calculating CPU resources by TKE Serverless with the frontend. (kube-scheduler)
      2021-12-02
      v1.20.6-tke.8
      Optimize grpc logs to avoid excessive logging when kubelet collects volume status. (kubelet)
      Avoid scheduling Pods using cbs to external CHC Nodes. (kube-scheduler)
      2021-11-26
      v1.20.6-tke.7
      Add support for custom installation of other CNI in hybrid cloud external nodes. (kube-controller-manager)
      Avoid unnecessary processing of updates after Pod is Assumed. (kube-scheduler)
      Merge pr99336 to improve the synchronization mechanism of node information when kubelet starts. (kubelet)
      2021-10-13
      v1.20.6-tke.6
      Merge 89465 to fix the issue of incorrect calculation of the number of instances by HPA based on pod metrics during rolling updates. (kube-controller-manager)
      2021-09-27
      v1.20.6-tke.5
      Support for collecting disk usage metrics for containerd runtime. (kubelet)
      2021-09-23
      v1.20.6-tke.4
      Fix the issue of no data for storage metrics when using cgroup v2. (kubelet)
      Fix CVE-2021-25741 to prevent illegal access to host files through symbolic links. (kubelet)
      2021-07-19
      v1.20.6-tke.3
      From TKE Cluster scaling to TKE Serverless, when batch scheduling Pods, it can correctly perceive the remaining IPs in the subnet and schedule the appropriate number of Pods to super nodes. (kube-scheduler)
      Port the upstream modifications for kubelet and cadvisor to fix the issue with metric collection statistics when using cgroupv2. (kubelet)
      2021-06-21
      v1.20.6-tke.2
      By default, CSIMigration and CSIMigrationQcloudCbs are enabled to mount CBS disks in CSI mode.
      2021-05-25
      v1.20.6-tke.1
      revert pr63066, fix the issue with LB health checks and IPVS. (kube-proxy)
      Merge pr90260 to fix the issue of missing network monitoring in containerd cluster. (kubelet)
      Fix the issue caused by the upgrade of lxcfs under ubuntu16 where Pods exit. (kubelet)
      Merge pr72914 to fix the issue where deleting a Pod and immediately creating and scheduling it to the same node might prevent it from being successfully mounted. (kube-controller-manager)
      Solve the issue where creating containers in CentOS leads to cgroup leaks. (kubelet)
      Merge pr98262 to support dynamic adjustment of log levels in kube-controller-manager. (kube-controller-manager)
      Merge pr97752 to fix the issue where NewReplicaSet is displayed when describing deployments. (kubectl)
      Merge pr94833 to fix the issue where the image tag in the status does not match when a Pod has multiple tags. (kubelet)
      Merge pr100060 and automatically delete directories of volumes left by orphaned Pods. (kubelet)
      kube-controller-manager supports super nodes. (kube-controller-manager)
      kube-scheduler supports retaining a fixed number of local replicas when scaling out to TKE Serverless in a hybrid cloud. (kube-scheduler)
      Support cbs csi migration. (kube-controller-manager, kubelet)
      Merge pr93260 to solve the problem of node startup slowing down caused by AWS Credential Provider. (kubelet)
      Add command-line parameter eks-config-namespace to the scheduler: Specify the namespace where the configuration related to scaling out TKE Serverless is located. (kube-scheduler)
      TKE supports hybrid cloud nodes. (kube-controller-manager)

      (Discontinued) TKE kubernetes 1.18.4 revisions

      Time
      Version
      Update Contents
      2024-04-10
      v1.18.4-tke.44
      Mounting PVC optimization. (kube-scheduler)
      2024-02-27
      v1.18.4-tke.43
      Super nodes allow scheduling Pods in TKE trunking ENI mode. (kube-scheduler)
      Optimization of computing resource policies for super nodes. (kube-scheduler)
      2024-02-01
      v1.18.4-tke.42
      Native Node Support for CVM Mode. (kubelet)
      Merge PR102014 to fix the issue that CPUs bound to init containers cannot be used after release. (kubelet)
      2023-12-04
      v1.18.4-tke.41
      CVE-2022-3294 (kube-apiserver)
      Read cache optimization. (kube-apiserver)
      Pods with added annotation support mounting cgroupfs. (kubelet)
      When deleting resources, add userAgent reporting and record detailed logs. (kube-apiserver)
      2023-07-20
      v1.18.4-tke.39
      Support scheduling daemonset pods to monthly subscription super nodes. (kube-scheduler)
      Port PR107831 to include pod UID in kubelet's secret/configmap cache. (kubelet)
      Fix pod mounting disk anomaly issue. (kube-controller-manager/kubelet)
      2023-05-26
      v1.18.4-tke.37
      Improve the processing logic for in-place Pod configuration changes. (kube-apiserver,kube-scheduler,kubelet)
      2023-05-16
      v1.18.4-tke.36
      When kube-controller-manager performs pod gc, prioritize deleting Evicted pods. (kube-controller-manager)
      Fix the issue where some GET requests were not logged as WATCH. (kube-apiserver)
      2023-03-03
      v1.18.4-tke.35
      Add delegated_to_storage tag to the list_too_many_objects_events_total metric, indicating whether the request is passed through to backend storage. (kube-apiserver)
      Port PR89160 to only delete the soft links related to a container after it has completely exited. (kubelet)
      Support exec/logs commands for super node DaemonSet Pod. (kube-apiserver)
      Optimize super node scheduling, add a unified switch EnableSuperNodeFeatures. (kube-scheduler)
      Fix inotify leaks causing kubelet memory leaks. (kubelet)
      Remove logs related to watch_too_old_objects_events_total, list_too_many_objects_events_total, watch_too_many_objects_events_total, to avoid printing excessive information. (kube-apiserver)
      Optimize self-adaptive node iptables mode Script, fixing misjudgments in some cases. (kube-proxy)
      Fix the issue where the metric rest_client_rate_limiter_duration_seconds was unregistered. (kube-apiserver, kube-controller-manager, kube-scheduler, kubelet, kube-proxy)
      Merge PR110951 to avoid the issue where sporadic PV mount operations get stuck, causing Pods associated with that PV to remain in the Creating state. (kubelet)
      Merge PR111635 to fix the issue where ipvs real server weight is 0 after kube-proxy restarts. (kube-proxy)
      2023-01-05
      v1.18.4-tke.34
      Super nodes support running DaemonSet Pods when needed. (kube-controller-manager)
      Merge PR102576 to no longer use RemoveAll when deleting legacy Pod volumes to avoid deleting volume-related data. (kubelet)
      Support lowering resource limits through in-place updates. (kube-apiserver)
      Fix the issue to not delete terminated containers during log rotation. (kubelet)
      Record metrics when list/watch is too old, too large, or there are too many objects (kube-apiserver): watch_too_old_objects_events_total, list_too_large_objects_events_total, list_too_many_objects_events_total, watch_too_many_objects_events_total
      2022-11-09
      v1.18.4-tke.33
      Optimize monthly subscription super node scheduling. (kube-scheduler)
      Optimize local replica count scheduling for super nodes.
      2022-10-13
      v1.18.4-tke.30
      Merge PR112299 to optimize the gzip compression level in apiserver. (kube-apiserver)
      Fix CVE-2022-3172. (kube-apiserver)
      Add calling_webhook_timeout_error metric. (kube-apiserver)
      Ignore TKEDefaultQuota conflict errors and increase worker count to resolve resource quota evaluates timeout errors. (kube-apiserver)
      Add StatusDetaching status for CBS disks. (kube-controller-manager)
      Serverless clusters support daemonset. (kube-apiserver)
      2022-09-07
      v1.18.4-tke.28
      Optimize List performance in large-scale clusters. (kube-apiserver)
      Optimize super node scheduling. (kube-scheduler)
      Support Pods under kube-system to be scheduled to prepaid super nodes. (kube-scheduler)
      Optimize super node HPA. (kube-controller-manager)
      Support in-place updates for Pod resources. (kube-apiserver, kubelet)
      Merge PR97348 to fix the issue where HPA scaling numbers are incorrect when StabilizationWindowSeconds is set. (kube-controller-manager)
      Optimize monthly subscription super node scheduling.
      Merge PR108831 to fix the issue where creating multiple Pods simultaneously causes kubelet to panic. (kubelet)
      2022-07-27
      v1.18.4-tke.26
      CBS supports native nodes. (kubelet)
      2022-03-18
      v1.18.4-tke.23
      Merge pr92878 to only print a warning message when setting ConfigMap/Secret volume Ownership takes more than 30 seconds, to avoid generating excessive log information. (kubelet)
      Merge pr106906 to check if the subpath of network storage volumes has been deleted, to prevent pods from being stuck in terminating status. (kubelet)
      When scheduling anti-affinity based on hostname, ignore TKE Serverless super nodes. (kube-scheduler)
      Merge pr93026 to fix the issue where DefaultPodTopologySpread cannot retrieve replicaset information. (kube-scheduler)
      2022-01-20
      v1.18.4-tke.20
      TKE Serverless Rescheduling Optimization: For the same availability zone's super nodes that have been evicted, lower their score. (kube-scheduler)
      apiserver supports integrating ExternalName 556 type external services. (kube-apiserver)
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      2021-12-09
      v1.18.4-tke.17
      Solve the issue where kube-controller-manager's access to api-server is rate-limited when there are a large number of volume attachment objects in the cluster. (kube-controller-manager)
      Merge PR95650, and HPA ignores deleted Pods when calculating the number of replicas. (kube-controller-manager)
      Fix the inconsistency in calculating CPU resources by TKE Serverless with the frontend. (kube-scheduler)
      2021-12-02
      v1.18.4-tke.16
      Fix the bug when scheduling to super nodes. (kube-scheduler)
      Optimize the super node scheduling algorithm. (kube-scheduler)
      2021-11-26
      v1.18.4-tke.15
      Merge pr96444 to return retry if there are errors while syncing RBAC policy. (kube-apiserver)
      Add support for custom installation of other CNI in hybrid cloud external nodes. (kube-controller-manager)
      Support cloud gaming Android container group core binding requirement. (kubelet)
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      Merge pr99336 to improve the synchronization mechanism of node information when kubelet starts. (kubelet)
      Repair CVE-2021-25741, prevent illegal access to host files through a symbolic link. (kubelet)
      Optimize the error information caused by cbs disk creation failure leading to scheduling timeout. (kube-scheduler)
      Optimize grpc logs to avoid excessive logging when kubelet collects volume status. (kubelet)
      Avoid scheduling Pods using cbs to external CHC Nodes. (kube-scheduler)
      2021-08-23
      v1.18.4-tke.14
      Scale-out from TKE Cluster to TKE Serverless: Support fixed IP. (kube-scheduler)
      Scale-out from TKE Cluster to TKE Serverless: When matching TKE Serverless fixed IP, skip other pre-selection policies. (kube-scheduler)
      From TKE Cluster scaling to TKE Serverless: Scheduling optimization for TKE Serverless nodes includes Node Resource-Aware Re-scheduling for TKE Serverless nodes; Priority Model Scheduling for TKE Serverless nodes; Optimization of Preferred/Pre-selected Policies for TKE Serverless nodes. (kube-scheduler)
      Record loaded ipvs kernel modules to avoid kube-proxy crashing in ipvs mode. (kube-proxy)
      Prevent panic when writing to the CPU manager status file fails. (kubelet)
      2021-07-22
      v1.18.4-tke.13
      Merge PR91859 to fix the issue where kube-apiserver panics when CRD types have only one letter. (kube-apiserver)
      2021-07-13
      v1.18.4-tke.12
      From TKE Cluster scaling to EKS: When batch scheduling Pods, it correctly perceives remaining Subnet IPs and schedules the appropriate number of Pods to super nodes. (kube-scheduler)
      Support for collecting disk usage metrics for Containerd runtime. (kubelet)
      Support specified Pods during scale-down. (kube-controller-manager)
      2021-06-05
      v1.18.4-tke.11
      TKE supports hybrid cloud nodes. (kube-controller-manager)
      2021-05-14
      v1.18.4-tke.9
      Port pr93370 to support CronJobControllerV2. (kube-controller-manager)
      Merge pr100376 to enable HTTP/2 health checks to avoid issues where connections cannot recover after being lost. (kube-apiserver, kube-controller-manager, kube-scheduler, kubelet, kube-proxy, kubectl)
      Merge pr100317 to fix CVE-2021-25735 where node updates could bypass the Validating Admission Webhook. (kube-apiserver)
      From TKE cluster scaling out to TKE Serverless support for ComputeResource and TKE Serverless ClusterIP, and HPA. (kube-controller-manager, kube-scheduler)
      2021-04-02
      v1.18.4-tke.8
      Merge pr97752 to fix the issue where NewReplicaSet appears as <none> when describing a deployment. (kubectl)
      Merge pr93808 to fix the issue of redundant information returned by executing kube-scheduler --version. (kube-scheduler)
      Merge pr91590 to fix the warning of ports already allocated when using NodePort type with multi-protocol Service. (kube-apiserver)
      Merge pr98262 to support dynamic adjustment of log levels in kube-controller-manager. (kube-controller-manager)
      Merge pr95154 to fix the issue of kube-scheduler snapshot including nodes that are in the process of being deleted. (kube-scheduler)
      Merge pr95711 to fix the issue of high CPU usage when executing the kubectl drain command. (kubectl)
      Merge pr96602 to fix the memory leaks in apiserver during timing jumps. (kube-apiserver)
      Merge pr97023 to delete related metadata directories when uninstalling an emptyDir type volume. (kubelet)
      Merge pr97527 to fix unsynchronized map access operations in cpumanager. (kubelet)
      Merge pr100190 to automatically delete orphaned Pod's leftover volume directories. (kubelet)
      Merge pr92614, and once all containers in a Pod with RestartPolicyOnFailure exit successfully, no new Sandbox will be created. (kubelet)
      Merge pr94833 to fix the issue where, if a Pod's image has multiple tags, the tags in the status do not match. (kubelet)
      2020-12-28
      v1.18.4-tke.6 (Starting from this version, ARM Clusters are supported)
      Add metrics for QcloudCbs. (kube-controller-manager)
      Fix the issue of superfluous spaces when checking the serial value while mounting a cbs disk. (kubelet)
      2020-12-21
      v1.18.4-tke.5
      Merge pr94712 to fix CVE-2020-8564 - Docker configuration leakage when the file format is incorrect and logLevel >= 4. (kubelet)
      Merge pr95316 to fix CVE-2020-8565 - Log token leakage due to an incomplete fix of CVE-2019-11250. (logLevel >= 9) (kube-apiserver, kubectl)
      Merge pr95245 to fix CVE-2020-8566 - Exposure of Ceph RBD adminSecrets in logs when logLevel >= 4. (kube-controller-manager)
      Fix the issue where restarting kubelet causes the Pod's readiness check to fail. (kubelet)
      Merge pr90825 to fix the issue where a race condition could cause the fifo queue Pop operation in client-go to get stuck, leading to pods remaining in a pending state indefinitely. (kubelet)
      The scheduler supports super nodes. (kube-scheduler)
      kube-controller-manager supports super nodes. (kube-controller-manager)
      Set the instance-type Tag based on the actual model of the node, no longer fixed to QCLOUD. (kubelet)
      Add the CBS section in OpenAPI. (kube-apiserver)
      Merge pr91126 to fix the issue where scheduler cache is inconsistent when Pods have the same name but different UIDs. (kube-scheduler)
      Merge pr93387 to fix the issue in the scheduler where disordered node cache information prevents daemonset pods from being scheduled on certain nodes. (kube-scheduler)
      Merge pr89465 to fix the issue of incorrect instance count calculation based on pod metrics during rolling updates. (kube-controller-manager)
      2020-10-13
      v1.18.4-tke.3
      Merge pr89629 to fix the issue where containers with subpath mounted from configmap never restart successfully after configmap changes. (kubelet)
      QcloudCbs supports BulkVolumeVerification. (kube-controller-manager)
      Merge pr94430 to fix the issue where client-go reflector couldn't detect the "Too large resource version" error. (kubelet)
      2020-08-12
      v1.18.4-tke.2
      Merge pr93403 to remove the erroneous print information from kubelet regarding the update of Pod Conditions that do not belong to kubelet. (kubelet)
      2020-08-04
      v1.18.4-tke.1
      Revert pr63066 to fix the issue with LB health checks and IPVS. (kube-proxy)
      Merge pr72914 to fix the issue where immediately creating and scheduling to the same node after deleting a Pod may result in unsuccessful mounting. (kube-controller-manager)
      Fix the issue where creating containers in CentOS leads to cgroup leaks. (kubelet)
      Fix for Pod exit issue caused by the upgrade of lxcfs under Ubuntu16. (kubelet)
      metadata increment Cache and Timed out. cloud-provider supports adding the node name as hostname. (kubelet)
      metadata adds local cache. (kubelet)
      Incorporate CBS and related fix code. (kubelet)
      Merge pr90260 to fix the missing Cluster Network Monitoring issue in containerd clusters. (kubelet)
      TKE supports the awareness of the maximum number of qcloudcbs that can be mounted on a single node. For versions 1.12 and above, it is maxAttachCount-2; for version 1.10, it is now defaulted to 18. (kube-scheduler)
      CBS intree resolves the issue of continuing to Uninstall disks when a disk does not exist, causing a large number of invalid requests. (kubelet)
      Merge pr2359 to fix the monitoring gap caused by being unable to obtain the docker root. (kubelet)
      kube-scheduler supports dynamic configuration of log levels. (kube-scheduler)
      Bypass the issue of missing device path (/dev/disk/by-id/virtio-xxx/...) encountered with CBS, allowing users to use CBS normally. (kubelet)
      TKE perceives the maximum number of qcloudcbs that can be mounted on a single node; kubelet does not patch the node. (kubelet)
      Merge pr89296, and no longer record whether iptables random-fully parameter is enabled. (kube-proxy)
      Fix aws issue, pr92162. (kubelet)
      Merge pr91277 to avoid the issue of CLB health checks leading to a large number of TLS handshake error logs in kube-apiserver. (kube-apiserver)
      Merge pr91500 to fix the issue of missing KUBERNETES_SERVICE_HOST environment variable. (kubelet)
      Merge pr92537 to fix the issue where client-go reflector couldn't recover from the "Too large resource version" error. (kube-apiserver, kube-controller-manager, kube-scheduler, kubelet, and kube-proxy)
      Merge pr92969 to fix the CVE-2020-8559 issue of escalating privileges from a compromised node to compromise other nodes. (kube-apiserver)
      Merge pr92921 to fix the CVE-2020-8557, a DOS attack issue by writing to "/etc/hosts" and exhausting disk space. (kubelet)

      (Discontinued) TKE kubernetes 1.16.3 revisions

      Time
      Version
      Update Contents
      2024-04-10
      v1.16.3-tke.40
      Native nodes support the cvm mode.(kubelet)
      Volume Reconciliation Optimization.(kube-controller-manager)
      Super nodes allow the scheduling of pods in TKE trunking ENI mode.(kube-scheduler)
      Scheduler Cache Issue Fixed.(kube-scheduler)
      Optimize PVC mounting. (kube-scheduler)
      2023-03-24
      v1.16.3-tke.35
      Optimize installation script. (kubelet)
      2023-03-03
      v1.16.3-tke.34
      Super nodes support running DaemonSet Pods when needed. (kube-controller-manager)
      Merge PR102576, and no longer use RemoveAll when deleting legacy Pod volumes to avoid deleting volume-related data. (kubelet)
      In-place update supports reducing resource limits. (kube-apiserver)
      Support exec/logs commands for super node DaemonSet Pod. (kube-apiserver)
      Fix inotify leaks causing kubelet memory leaks. (kubelet)
      Optimize self-adaptive node iptables mode Script, fixing misjudgments in some cases. (kube-proxy)
      Merge PR111635 to fix the issue where ipvs real server weight is 0 after kube-proxy restarts. (kube-proxy)
      2022-10-13
      v1.16.3-tke.32
      Merge PR112299 to optimize the gzip compression level in apiserver. (kube-apiserver)
      Fix CVE-2022-3172. (kube-apiserver)
      Add calling_webhook_timeout_error metric. (kube-apiserver)
      Ignore TKEDefaultQuota conflict errors, and add worker count to resolve resource quota evaluates timeout errors. (kube-apiserver)
      CPU Manager cleans up terminated containers before processing Pods, to avoid CPU allocation failure. (kubelet)
      Add StatusDetaching state for CBS disks. (kube-controller-manager)
      Serverless clusters support daemonset. (kube-apiserver)
      Support in-place updates for Pod resources. (kube-apiserver, kubelet)
      2022-07-27
      v1.16.3-tke.28
      When scheduling based on hostname anti-affinity, ignore TKE Serverless virtual nodes. (kube-scheduler)
      TKE Serverless supports retaining sandbox features. (kube-scheduler)
      CBS supports native nodes. (kubelet)
      2022-03-18
      v1.16.3-tke.27
      Support specified Pods during scale-down. (kube-controller-manager)
      Optimize the super node scheduling algorithm. (kube-scheduler)
      2022-01-20
      v1.16.3-tke.25
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      The apiserver supports integration with ExternalName type external services. (kube-apiserver)
      Optimize TKE Serverless scheduling. (kube-scheduler)
      2021-12-09
      v1.16.3-tke.24
      Fix the issue where the TKE Serverless local replica count policy does not work on statefulset type Pods. (kube-scheduler)
      2021-12-02
      v1.16.3-tke.23
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      Avoid scheduling Pods using cbs to external CHC Nodes. (kube-scheduler)
      Fix the bug when scheduling to super nodes. (kube-scheduler)
      2021-09-03
      v1.16.3-tke.22
      Avoid panic when errors occur when writing the cpu manager status file. (kubelet)
      2021-08-17
      v1.16.3-tke.21
      Optimizations for scheduling TKE Serverless nodes: Node Resource-Aware Re-scheduling for TKE Serverless nodes; Priority Model Scheduling for TKE Serverless nodes; Optimized Optimal Selection / Pre-selection Policy for TKE Serverless nodes. (kube-scheduler)
      Port 87692 to fix the issue where the scheduler's pending_pods and schedule_attempts_total metrics have no data. (kube-scheduler)
      2021-07-19
      v1.16.3-tke.20
      Port 87688 and 87693 to optimize Node Authorizer performance. (kube-apiserver)
      Scaling from TKE clusters to TKE Serverless: Accurately perceive the remaining IPs of the subnet when batch scheduling Pods, scheduling the correct number of Pods to super nodes. (kube-scheduler)
      Merge pr88507 to fix the issue where podIP and podIPs are inconsistent when updating Pod status. (kube-apiserver)
      2021-05-24
      v1.16.3-tke.17
      Port pr93370 to support CronJobControllerV2. (kube-controller-manager)
      Scaling from TKE clusters to TKE Serverless supports retaining the local replica count. (kube-scheduler)
      2021-05-06
      v1.16.3-tke.16
      Update the startup mode when running kube-proxy in image method to automatically adapt to the iptables running mode of the node, to support operating systems that by default use the nf_tables mode for iptables.
      2021-04-14
      v1.16.3-tke.15
      Merge pr97752 to fix the issue where NewReplicaSet appears as <none> when describing a deployment. (kubectl)
      Merge pr92614, when the restart policy is RestartPolicyOnFailure and all containers in a Pod exit successfully, a new Sandbox is no longer created. (kubelet)
      Merge pr91590 to fix the issue of port already allocated warnings when using NodePort type services with multiple protocols. (kube-apiserver)
      Merge pr98262 to support for dynamic log level adjustment in kube-controller-manager. (kube-controller-manager)
      Merge pr95301 to automatically delete the volume directories left by orphan Pods. (kubelet)
      2020-12-28
      v1.16.3-tke.14
      Add metrics for QcloudCbs. (kube-controller-manager)
      Fix the issue of extraneous spaces when viewing the serial value during mount of cbs disks. (kubelet)
      2020-12-21
      v1.16.3-tke.13
      Merge pr94712 to fix CVE-2020-8564 - Docker configuration leakage when the file format is incorrect and logLevel >= 4. (kubelet)
      Merge pr95316 to fix CVE-2020-8565 - Log token leakage due to incomplete fix of CVE-2019-11250 (logLevel >= 9). (kube-apiserver, kubectl)
      Merge pr95245 to fix CVE-2020-8566 - Exposure of Ceph RBD adminSecrets in logs when logLevel >= 4. (kube-controller-manager)
      Merge pr86191 to fix the issue where pods may be in an Error Status after a node restart. (kubelet)
      Merge pr86140 to fix the issue where the controller manager fails to properly handle Timeout Errors, preventing the creation of Scale-out pods. (kube-controller-manager)
      Merge pr90825 to address the issue where FIFO Queue Pop Operations in client-go might get stuck due to race conditions, causing pods to remain in a Pending State indefinitely. (kubelet)
      The scheduler supports super nodes. (kube-scheduler)
      kube-controller-manager supports super nodes. (kube-controller-manager)
      Set the instance-type tag based on the actual model of the node, no longer fixed to QCLOUD. (kubelet)
      Add the CBS section in OpenAPI. (kube-apiserver)
      Merge pr81344 to fix the issue where CPU Manager does not support SourcesReady. (kubelet)
      Merge pr91126 to fix the inconsistency in scheduler cache when Pods have the same name but different UIDs. (kube-scheduler)
      Merge pr89224 to fix the issue where the absence of a check in NodeInfo caused kube-scheduler to crash unexpectedly. (kube-scheduler)
      Merge pr89465 to fix the issue where HPA calculates the wrong number of instances during rolling updates based on pod metrics. (kube-controller-manager)
      2020-10-13
      v1.16.3-tke.11
      Merge pr92971 to fix CVE-2020-8559 issue of escalating privileges from a compromised node to invade other nodes. (kube-apiserver)
      Merge pr92924 to fix CVE-2020-8557 DOS attack issue by writing to /etc/hosts to deplete disk space. (kubelet)
      Merge pr93403 to remove erroneous print information in kubelet for updating Pod Conditions not owned by kubelet. (kubelet)
      Merge pr89629 to resolve the issue where the container restart always fails after configmap changes when mounting subpath. (kubelet)
      QcloudCbs supports BulkVolumeVerification. (kube-controller-manager)
      Merge pr84998 to fix the issue where the corresponding node lease objects might be recreated after a node is deleted, leading to garbage data. (kubelet)
      2020-07-28
      v1.16.3-tke.10
      Merge pr91277 to fix the issue where CLB health checks cause kube-apiserver to generate a large number of TLS handshake error logs. (kube-apiserver)
      Merge pr91500 to fix the issue of missing KUBERNETES_SERVICE_HOST environment variable. (kubelet)
      2020-06-17
      v1.16.3-tke.9
      Temporary fix for an AWS issue pr92162. Stops registering AWS Credential Provider to avoid the node startup slowdown it causes.
      2020-06-11
      v1.16.3-tke.8
      Merge pr85993 to support setting the gateway address for kubenet using CNI results.
      2020-06-10
      v1.16.3-tke.7
      Merge pr90260 to fix the issue of missing network monitoring for containerd cluster.
      Merge pr89515 to fix the issue where HPA calculates the wrong number of instances during rolling updates.
      Merge pr91252 to ignore Pod Condition updates from other components to avoid unnecessary scheduling.
      Merge pr89794 to clean up the error logs of kube-controller-manager to prevent Half-Blind SSRF attacks in CVE-2020-8555.
      2020-05-18
      v1.16.3-tke.6
      TKE can sense the maximum number of qcloudcbs mountable on a single node, does not support dynamically acquiring the maximum value.
      2020-04-20
      v1.16.3-tke.5
      Merge pr69047 to solve the backward compatibility issue with node.Spec.Unschedulable (this fix was overridden when merging in-tree CBS code).
      2020-04-14
      v1.16.3-tke.4
      Merge pr87913 to fix CVE-2020-8551: Kubelet DoS attack issue.
      Merge pr87669 to fix CVE-2020-8552: apiserver DoS attack issue.
      TKE supports sensing the maximum quantity of qcloudcbs that can be mounted on a single node (For version 1.12 and above, it's maxAttachCount-2; for version 1.10, the default is currently 18).
      Merge pr87467 to fix the issue of authorized users sending malicious YAML causing kubectl to consume excessive CPU during YAML parsing.
      2020-03-11
      v1.16.3-tke.3
      cbs intree fixes the issue of continuing to uninstall disks when the disk does not exist, leading to a large number of invalid requests.
      Metadata adds local caching.
      2020-02-14
      v1.16.3-tke.2
      Merge pr2359 to solve the monitoring missing issue caused by the inability to obtain docker root.
      Merge pr86583 to improve the log output level when iptables does not support random-fully, avoiding excessive log generation.
      kube-scheduler supports dynamic setting of log levels.
      Bypass the missing device path (/dev/disk/by-id/virtio-xxx/...) issue with CBS, allowing users to use CBS normally.
      Merge pr86230 to skip updating the assumed pod's scheduling during the pod scheduling process.
      2020-01-06
      v1.16.3-tke.1
      Merge pr79036 to address the issue where, when the cpumanager is enabled, if the pod's QoS is Guaranteed, then the CPU quota is disabled.
      Merge pr84167 to fix the issue of the apiserver health check failing due to incorrect ETCD key prefix.
      Revert pr63066 to fix the issues with CLB health checks and IPVS.
      Merge pr72914 to fix the issue where immediately creating and scheduling a Pod to the same node after deletion may lead to unsuccessful mounting.
      Fix the issue of cgroup leakage caused by creating containers under CentOS.
      Fix the issue of pod exit caused by lxcfs upgrade under Ubuntu16.
      metadata increases cache and timeout, cloud-provider adds support for using node name as hostname.
      Revert pr79036 to address the issue where, when the cpumanager is enabled, if the pod's QoS is Guaranteed, the CPU quota is disabled.
      Bypass the missing device path (/dev/disk/by-id/virtio-xxx/...) issue with CBS, allowing users to use CBS normally.

      (Discontinued) TKE kubernetes 1.14.3 revisions

      Time
      Version
      Update Contents
      2024-04-10
      v1.14.3-tke.31
      Volume Reconciliation Optimization.(kube-controller-manager)
      Super nodes allow the scheduling of pods in TKE trunking ENI mode.(kube-scheduler)
      2023-03-03
      v1.14.3-tke.28
      Added StatusDetaching state for CBS disks. (kube-controller-manager)
      Fixed CVE-2022-3172. (kube-apiserver)
      Port 16+ webhook metrics to add calling_webhook_timeout_error metric. (kube-apiserver)
      Port etcd_request_duration_seconds metric. (kube-apiserver)
      Optimize self-adaptive node iptables mode Script to fix misjudgments in some cases. (kube-proxy)
      Merge PR111635 to fix the issue where ipvs real server weight is 0 after kube-proxy restarts. (kube-proxy)
      2022-04-13
      v1.14.3-tke.27
      Merge pr78428, avoid writing warning messages when exporting iptables regulations, to prevent kube-proxy panic during recovery. (kube-proxy)
      2022-03-18
      v1.14.3-tke.26
      Support specified Pods during scale-down. (kube-controller-manager)
      Optimize the super node scheduling algorithm. (kube-scheduler)
      Merge pr80851 to fix CVE-2019-11247, prevent unauthorized access to CRD resources. (kube-apiserver)
      2022-01-20
      v1.14.3-tke.24
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      The apiserver supports integration with ExternalName type external services. (kube-apiserver)
      Optimize TKE Serverless scheduling. (kube-scheduler)
      2021-12-02
      v1.14.3-tke.23
      Scaling from TKE clusters to TKE Serverless: Accurately perceive the remaining IPs of the subnet when batch scheduling Pods, scheduling the correct number of Pods to super nodes. (kube-scheduler)
      TKE Serverless Node Scheduling Optimization: Resource-aware re-scheduling for TKE Serverless nodes; Priority model scheduling for TKE Serverless nodes; Optimized preferred/pre-selected policies for TKE Serverless nodes. (kube-scheduler)
      Support extended scheduler Prebind and Unreserve operations. (kube-scheduler)
      Avoid scheduling Pods using cbs to external CHC Nodes. (kube-scheduler)
      Fix the bug when scheduling to super nodes. (kube-scheduler)
      2021-05-06
      v1.14.3-tke.22
      Update the startup mode for running kube-proxy as an image, to automatically adapt to the iptables operating mode of the node, supporting operating systems that default to using the nf_tables mode for iptables.
      2021-04-14
      v1.14.3-tke.21
      Merge pr97752 to fix the issue where NewReplicaSet is displayed as <none> when describing a deployment. (kubectl)
      Merge pr78999 to fix the case sensitivity issue in protocol determination during graceful shutdown. (kube-proxy)
      Merge pr91590 to fix the issue of port already allocated warnings when using NodePort type services with multiple protocols. (kube-apiserver)
      Merge pr98262 to support for dynamic log level adjustment in kube-controller-manager. (kube-controller-manager)
      Merge pr95301 to automatically delete the volume directories left by orphan Pods. (kubelet)
      2020-12-28
      v1.14.3-tke.19
      Add metrics for QcloudCbs. (kube-controller-manager)
      Fix the issue of extraneous spaces when viewing the serial value during mount of cbs disks. (kubelet)
      2020-12-21
      v1.14.3-tke.18
      Merge pr94712 to fix CVE-2020-8564 - Docker configuration leakage when the file format is incorrect and logLevel >= 4. (kubelet)
      Merge pr95316 to fix CVE-2020-8565 - Log token leakage due to an incomplete fix of CVE-2019-11250 (logLevel >= 9). (kube-apiserver, kubectl)
      Merge pr95245 to fix CVE-2020-8566 - Exposure of Ceph RBD adminSecrets in logs when logLevel >= 4. (kube-controller-manager)
      Merge pr86140 to fix the issue where the controller manager fails to properly handle timeout errors, preventing the creation of scale-out pods. (kube-controller-manager)
      The scheduler supports super nodes. (kube-scheduler)
      kube-controller-manager supports super nodes. (kube-controller-manager)
      Set the instance-type Tag based on the actual model of the node, no longer fixed to QCLOUD. (kubelet)
      Merge pr79338 to not enable the pids cgroup subsystem when neither SupportPodPidsLimit nor SupportNodePidsLimit is enabled. (kubelet)
      Merge pr89224 to fix the issue where the absence of a check in NodeInfo caused kube-scheduler to crash unexpectedly. (kube-scheduler)
      Merge pr89465 to fix the issue where HPA calculates the wrong number of instances during rolling updates based on pod metrics. (kube-controller-manager)
      2020-10-13
      v1.14.3-tke.17
      Merge pr74781 to change the default update policy of ConfigMap and Secret from Cache to Watch. (kubelet)
      Merge pr93403 to remove erroneous print information in kubelet for updating Pod Conditions not owned by kubelet. (kubelet)
      Merge pr89629 to resolve the issue where the container restart always fails after configmap changes when mounting subpath. (kubelet)
      Merge pr80942 to fix the issue where rules are not deleted after a service is deleted in ipvs mode. (kube-proxy)
      QcloudCbs supports BulkVolumeVerification. (kube-controller-manager)
      2020-08-04
      v1.14.3-tke.16
      Merge pr78883 to fix the bug where a default value is automatically added to pod.spec.container.SecurityContext.ProcMount.
      2020-07-28
      v1.14.3-tke.15
      Merge pr76518 and pr82514 to limit the return size of HTTP and exec probes to prevent the consumption of a significant amount of node memory. (kubelet)
      Merge pr91277 to avoid issues where CLB health checks cause kube-apiserver to generate a large number of TLS handshake error logs. (kube-apiserver)
      Merge pr91500 to fix the issue of missing KUBERNETES_SERVICE_HOST environment variable. (kubelet)
      Merge pr77475 to fix the issue where Cronjob cannot schedule when the number of Jobs exceeds 500. (kube-controller-manager)
      2020-06-10
      v1.14.3-tke.14
      Merge pr85027 to fix the issue where HPA calculates the wrong number of instances during rolling updates.
      Merge pr79708 to use spec.replicas to calculate the current number of HPA replicas.
      Merge pr91252 to ignore Pod Condition updates from other components to avoid unnecessary scheduling.
      Merge pr89794 to clean up the error logs of kube-controller-manager to prevent Half-Blind SSRF attacks in CVE-2020-8555.
      2020-06-04
      v1.14.3-tke.13
      Merge pr90260 to fix the issue of missing network monitoring for containerd cluster.
      Merge pr79451 to fix the issue of missing cluster not recreate a Sandbox after a failure when restartPolicy is set to Never.
      2020-05-18
      v1.14.3-tke.12
      TKE can perceive the maximum number of qcloudcbs that can be mounted on a single node, but does not support dynamically obtaining the maximum value.
      2020-04-14
      v1.14.3-tke.11
      Merge pr75442 to correct the bandwidth unit from Kb to b.
      Merge pr87669 to fix CVE-2020-8552: apiserver DoS attack issue.
      TKE supports perceiving the maximum quantity of qcloudcbs that can be mounted on a single node (For version 1.12 and above, it's maxAttachCount-2; for version 1.10, the current default is 18).
      2020-04-14
      v1.14.3-tke.10
      CBS intree fixes the issue of a large number of invalid requests caused by continuing to Uninstall a disk when the disk does not exist.
      2020-01-13
      v1.14.3-tke.9
      Merge pr2359 to address the monitoring deficiency issues caused by the inability to obtain the docker root.
      Merge pr86583 to improve the log output level when iptables does not support random-fully, avoiding excessive log generation.
      kube-scheduler supports dynamic setting of log levels.
      Bypass the missing device path (/dev/disk/by-id/virtio-xxx/...) issue with CBS, allowing users to use CBS normally.
      Merge pr86230 to skip updating the assumed pod's scheduling during the pod scheduling process.
      2019-12-23
      v1.14.3-tke.8
      Revert pr79036 to resolve the issue where, when cpumanager is enabled, the cpu quota is turned off if the Pod's QoS is Guaranteed.
      2019-12-17
      v1.14.3-tke.7
      Metadata enhancements include cache and timeout.
      Fix the issue where upgrading lxcfs under Ubuntu16 causes Pods to exit.
      Restart kubelet to avoid pods not ready due to readiness.
      2019-11-28
      v1.14.3-tke.6
      cloud-provider adds support for using the node name as hostname.
      2019-11-18
      v1.14.3-tke.5
      Merge pr83435 to fix the issue where an attacker could send specially crafted malicious YAML or JSON attack payloads causing kube-apiserver CPU or memory exhaustion and inability to provide services.
      Merge pr84167 to fix the issue where an incorrect Etcd key prefix causes the apiserver health check to fail.
      Merge pr75622 to fix the issue of delayed synchronization of changes to sts to pods (about 20s) in clusters with a large number of sts (>2000) workloads.
      2019-10-23
      v1.14.3-tke.4
      Merge pr79036 to fix the issue of turning off the cpu quota when CPU Manager is enabled and the Pod's QoS is Guaranteed.
      2019-09-10
      v1.14.3-tke.3
      Merge pr63066 to fix the issue of health check failure in IPVS mode for CLB.
      2019-09-06
      v1.14.3-tke.2
      Fix cve-2019-9512&cve-2019-9514 HTTP/2 DDoS Security Vulnerabilities.
      Merge pr72914 to fix the issue that deleting a Pod and then immediately creating and scheduling it to the same node may cause a volume mount to fail.
      Fix the issue of cgroup leakage caused by creating containers under CentOS.

      (Discontinued) TKE kubernetes 1.12.4 revisions

      Time
      Version
      Update Contents
      2023-03-03
      v1.12.4-tke.32
      Fix CVE-2022-3172. (kube-apiserver)
      Port 16+ webhook metrics to add calling_webhook_timeout_error metric. (kube-apiserver)
      Port etcd_request_duration_seconds metric. (kube-apiserver)
      Optimize self-adaptive node iptables mode Script to fix misjudgments in some cases. (kube-proxy)
      Merge PR111635 to fix the issue where ipvs real server weight is 0 after kube-proxy restarts. (kube-proxy)
      2022-04-13
      v1.12.4-tke.31
      Merge pr78428 to avoid writing warning messages when exporting iptables rules, causing kube-proxy to panic during recovery. (kube-proxy)
      2022-01-20
      v1.12.4-tke.30
      Support binding LB addresses to IPVS NICs. (kube-proxy)
      2021-05-06
      v1.12.4-tke.28
      Update the startup mode of kube-proxy when running as a container, automatically adapting to the iptables running mode of the node, to support operating systems that default to running iptables in nf_tables mode.
      2020-12-28
      v1.12.4-tke.27
      Add metrics for QcloudCbs. (kube-controller-manager)
      Fix the issue of superfluous spaces when checking the serial value while mounting a cbs disk. (kubelet)
      2020-12-15
      v1.12.4-tke.26
      QcloudCbs supports BulkVolumeVerification. (kube-controller-manager)
      2020-11-17
      v1.12.4-tke.25
      Merge pr79495 to fix the issue of webhook call failures when CRD has multiple versions. (kube-apiserver)
      2020-10-13
      v1.12.4-tke.24
      Merge pr93403 to remove the erroneous print information from kubelet regarding the update of Pod Conditions that do not belong to kubelet. (kubelet)
      2020-08-04
      v1.12.4-tke.23
      Merge pr78881 to fix the bug that by default adds a default value to pod.spec.container.SecurityContext.ProcMount.
      2020-07-28
      v1.12.4-tke.22
      Merge pr91277 to avoid the issue of CLB health checks leading to a large number of TLS handshake error logs in kube-apiserver. (kube-apiserver)
      Merge pr91500 to fix the issue of missing KUBERNETES_SERVICE_HOST environment variable. (kubelet)
      2020-06-10
      v1.12.4-tke.21
      Merge pr73915 to avoid watcher receiving events before the start of watch.
      Merge pr91252 to ignore Pod Condition updates produced by other components to avoid unnecessary scheduling.
      Merge pr73915 to clean up the error logs of kube-controller-manager to avoid Half-Blind SSRF attacks in CVE-2020-8555.
      2020-06-04
      v1.12.4-tke.20
      Merge pr90260 to fix the missing Cluster Network Monitoring issue in containerd.
      Merge pr79451 to fix the issue where kubelet fails to recreate SandBox after failure when restartPolicy is set to Never.
      2020-05-18
      v1.12.4-tke.19
      Merge pr77802 to disable graceful termination for UDP traffic.
      Merge pr68741 to fix the issue with symbolic links to /var/lib/kubelet and the use of subpaths causing mount point leakage and pods remaining in terminating state after deletion.
      TKE can detect the maximum number of qcloudcbs that can be mounted on a single node, and does not support dynamically acquiring the maximum value.
      2020-04-14
      v1.12.4-tke.18
      Merge pr73401, pr73606, and pr76060 to delete DaemonSet Pods assigned to non-existent nodes.
      Merge pr68619 to fix the issue of the cpumanager dirty data.
      Merge pr87669 to fix CVE-2020-8552: apiserver DoS attack issue.
      TKE supports detecting the maximum number of qcloudcbs that a single node can mount (for version 1.12 and above, it is maxAttachCount-2; for version 1.10, the default is currently 18).
      2020-02-14
      v1.12.4-tke.17
      Upgrade the CBS V2 interface to V3.
      CBS intree resolves the issue of a large number of invalid requests caused by continuing to Uninstall a disk when the disk does not exist.
      2020-01-13
      v1.12.4-tke.16
      Merge pr2359 to fix the issue of missing monitoring due to the inability to obtain the docker root.
      Merge pr86583 to improve the log output level when iptables does not support random-fully, avoiding excessive log generation.
      kube-scheduler supports dynamic setting of log levels.
      Bypass the issue where the device path (/dev/disk/by-id/virtio-xxx/...) is missing for cbs, allowing users to use cbs normally.
      Merge pr86230 to skip the update of the assumed pod's scheduling during the pod scheduling process.
      2019-12-23
      v1.12.4-tke.15
      Revert pr79036 to fix the issue where, when cpumanager is enabled, the cpu quota is turned off if the Pod's QoS is Guaranteed.
      2019-12-17
      v1.12.4-tke.14
      Metadata enhancements include cache and timeout.
      Fix the issue where upgrading lxcfs under Ubuntu16 causes Pods to exit.
      Restart kubelet to avoid pods not ready due to readiness.
      2019-11-28
      v1.12.4-tke.13
      cloud-provider adds support for using the node name as hostname.
      2019-11-18
      v1.12.4-tke.12
      Merge pr75622 to fix the issue of delayed synchronization of changes to sts to pods (about 20s) in clusters with a large number of sts (>2000) workloads.
      2019-10-23
      v1.12.4-tke.11
      Merge pr79036 to fix the issue where and when CPU Manager is enabled, the CPU quota is turned off if the Pod's QoS is Guaranteed.
      Merge pr72866 to add a --metrics-port command-line parameter to kube-proxy, and solve the bug where --metrics-bind-address could not include port.
      2019-09-06
      v1.12.4-tke.10
      Address the cve-2019-9512&cve-2019-9514 HTTP/2 DDoS security vulnerability.
      Merge pr72914 to fix the issue where deleting a Pod and immediately creating and scheduling it to the same node might cause a failure to mount the volume.
      Merge pr71834 to fix the issue where, in IPVS mode, sessionAffinity set to ClientIP could cause access failure to the RS.
      2019-08-09
      v1.12.4-tke.9
      Fix the issue where creating containers in CentOS leads to cgroup leaks.
      2019-08-08
      v1.12.4-tke.8
      Merge pr72118 to fix the issue where StatefulSets based on CBS cannot be mounted when rescheduled to the same node.
      2019-07-17
      v1.12.4-tke.7
      Merge pr75037 to address the security vulnerabilities in the kubectl cp command.
      2019-07-16
      v1.12.4-tke.6
      Resolve compatibility issues between the tlinux kernel version and IPVS, and fix the issue where health checks fail in IPVS mode for CLB.
      2019-07-09
      v1.12.4-tke.5
      Merge pr72361 to fix the possible Deadlock issue in kube-proxy.
      2019-06-25
      v1.12.4-tke.4
      Resolve compatibility issues between the tlinux kernel version and IPVS.
      2019-06-17
      v1.12.4-tke.3
      Merge pr71114 to fix throughput issues with IPVS.
      2019-06-04
      v1.12.4-tke.2
      Merge pr74755 to fix the kubelet hang issue.
      Merge pr69047 to fix the backward compatibility issue with node.Spec.Unschedulable.

      (Discontinued) TKE kubernetes 1.10.5 revisions

      Time
      Version
      Update Contents
      2021-05-06
      v1.10.5-tke.20
      Update the startup mode of kube-proxy when running as a container, automatically adapting to the iptables running mode of the node, to support operating systems that default to running iptables in nf_tables mode.
      2020-06-10
      v1.10.5-tke.19
      Merge pr90260 to fix the missing Cluster Network Monitoring issue in containerd.
      Merge pr91252 to ignore Pod Condition updates produced by other components to avoid unnecessary scheduling.
      Merge pr89794 to clean up the error logs of kube-controller-manager to prevent Half-Blind SSRF attacks in CVE-2020-8555.
      2020-05-18
      v1.12.4-tke.19
      Merge pr61549, add volumeSpec data to the mountedPods cache to resolve issues with multiple pods using the same volume not being able to be deleted normally.
      2020-04-29
      v1.10.5-tke.17
      Merge pr75622 to fix the issue of significant latency (~20s) in syncing sts to Pods in clusters with a large number of sts (>2000) workloads.
      2020-04-14
      v1.10.5-tke.16
      Merge pr68619 to fix the cpumanager dirty data issues.
      Merge pr87669 to fix CVE-2020-8552: apiserver DoS attack issue.
      TKE supports detecting the maximum number of qcloudcbs that a single node can mount (for version 1.12 and above, it is maxAttachCount-2; for version 1.10, the default is currently 18).
      2020-02-14
      v1.10.5-tke.15
      Upgrade the CBS V2 interface to V3.
      CBS intree resolves the issue of a large number of invalid requests caused by continuing to uninstall a disk when the disk does not exist.
      2020-01-13
      v1.10.5-tke.14
      Merge pr2359 to fix the issue of missing monitoring due to the inability to obtain the docker root.
      Merge pr86583 to improve the log output level when iptables does not support random-fully, avoiding excessive log generation.
      kube-scheduler supports dynamic setting of log levels.
      Bypass the issue where the device path (/dev/disk/by-id/virtio-xxx/...) is missing for cbs, allowing users to use cbs normally.
      Merge pr86230 to skip the update of the assumed pod's scheduling during the pod scheduling process.
      2019-12-23
      v1.10.5-tke.13
      Revert pr79036 to fix the issue where, when cpumanager is enabled, the cpu quota is turned off if the Pod's QoS is Guaranteed.
      2019-12-13
      v1.10.5-tke.12
      kubelet does not delete node when checking externalID.
      Metadata enhancements include cache and timeout.
      Fix the issue where upgrading lxcfs under Ubuntu16 causes Pods to exit.
      Restart kubelet to avoid pods not ready due to readiness.
      2019-11-18
      v1.10.5-tke.11
      Remove reverse detection from kube-controller-manager.
      2019-10-23
      v1.10.5-tke.10
      Merge pr79036 to fix the issue where and when CPU Manager is enabled, the CPU quota is turned off if the Pod's QoS is Guaranteed.
      Merge pr72866 to add a --metrics-port command-line parameter to kube-proxy, and solve the bug where --metrics-bind-address could not include port.
      2019-09-06
      v1.10.5-tke.9
      Address the cve-2019-9512&cve-2019-9514 HTTP/2 DDoS security vulnerability.
      Merge pr72914 to fix the issue where deleting a Pod and immediately creating and scheduling it to the same node might cause a failure to mount the volume.
      Merge 67430 to fix the data structure rollback in the event of a failure in updateContainerCPUSet.
      2019-08-08
      v1.10.5-tke.8
      Merge pr72118 to fix the issue where kubelet reports "resource name may not be empty" when attempting to Mount the same device immediately after an Unmount.
      2019-07-17
      v1.10.5-tke.7
      Merge pr75037 to fix the security vulnerabilities in the kubectl cp command.
      2019-06-25
      v1.10.5-tke.6
      Resolve compatibility issues between the tlinux kernel version and IPVS.
      2019-06-17
      v1.10.5-tke.5
      Merge pr71114 to fix throughput issues with IPVS.
      2019-03-19
      v1.10.5-tke.4
      Merge pr65092 to fix the panic issue in apiserver when handling specific requests.
      2019-02-19
      v1.10.5-tke.3
      Merge pr67288 to fix the connection leakage issue when apiserver acts as a proxy.
      2018-09-28
      v1.10.5-tke.2
      Move the logic of creating CLB out of controller-manager (implemented through a separate service controller).
      2018-09-27
      v1.10.5-tke.1
      backport pr63321 to fix the issue of excessively long Terminating time when there are multiple business containers in a pod.
      2018-09-21
      v1.10.5-qcloud-rev1
      When kubelet update status times out, controller-manager probes the kubelet port.

      (Discontinued) TKE kubernetes 1.8.13 revisions

      Time
      Version
      Update Contents
      2020-01-13
      v1.8.13-tke.7
      Merge pr2359 to fix the issue of missing monitoring due to the inability to obtain the docker root.
      Bypass the issue where the device path (/dev/disk/by-id/virtio-xxx/...) is missing for cbs, allowing users to use cbs normally.
      2019-12-13
      v1.8.13-tke.6
      kubelet does not delete node when checking externalID.
      Metadata enhancements include cache and timeout.
      Fix the issue where upgrading lxcfs under Ubuntu16 causes Pods to exit.
      Restart kubelet to avoid pods not ready due to readiness.
      2019-11-18
      v1.8.13-tke.5
      Remove reverse detection from kube-controller-manager.
      Add metric to cbs pvc.
      2018-09-28
      v1.8.13-tke.2
      Move the logic of creating CLB out of controller-manager (implemented through a separate service controller).
      2018-09-27
      v1.8.13-tke.1
      Disable kmem statistics to prevent leakage of cgroup quantities.
      Reduce resourcequota conflicts when creating pods.
      2018-09-21
      v1.8.13-qcloud-rev1
      When kubelet update status times out, controller-manager probes the kubelet port.

      (Discontinued) TKE kubernetes 1.7.8 revisions

      Time
      Version
      Update Contents
      2019-12-17
      v1.7.8-tke.4
      kubelet does not delete node when checking externalID.
      Metadata enhancements include cache and timeout.
      Fix the issue where upgrading lxcfs under Ubuntu16 causes Pods to exit.
      Restart kubelet to avoid pods not ready due to readiness.
      2018-09-28
      v1.7.8-tke.2
      Resolve the conflict issue between controller-manager and external service controller.
      2018-09-27
      v1.7.8-tke.1
      Move the logic of creating CLB out of controller-manager (implemented through a separate service controller).
      2018-09-21
      v1.7.8-qcloud-rev1
      When kubelet update status times out, controller-manager probes the kubelet port.
      
      
      
      
      
      Contact Us

      Contact our sales team or business advisors to help your business.

      Contact Us
      Technical Support

      Open a ticket if you're looking for further assistance. Our Ticket is 7x24 avaliable.

      Submit a Ticket
      7x24 Phone Support
      Copyright © 2013-2024 Tencent Cloud. All Rights Reserved.
      Privacy PolicyLegalCookie Policy