- Kubeadm: support automatic retry after failing to pull image ([#86899](https://github.com/kubernetes/kubernetes/pull/86899), [@SataQiu](https://github.com/SataQiu)) [SIG Cluster Lifecycle]

* Default resourceGroup should be used when the value of annotation azure-load-balancer-resource-group is an empty string. ([#79514](https://github.com/kubernetes/kubernetes/pull/79514), [@feiskyer](https://github.com/feiskyer))
* fix: retry detach azure disk issue ([#78700](https://github.com/kubernetes/kubernetes/pull/78700), [@andyzhangx](https://github.com/andyzhangx))
    * try to only update vm if detach a non-existing disk when got <200, error> after detach disk operation

      // 408's are a bit of an outlier because we may repeat the request if we encounter this
      // response code. In this scenario, there are 2 responses: the initial 408 and then the 200
      // because of the retry. We just want to ensure the initial 408 isn't cached.
      expectedResponseCode = 200
      server.enqueue(
        MockResponse.Builder()
          .setHeader("Cache-Control", "no-store")
          .body("FGHIJ")

* fix incorrect error info when creating an azure file PVC failed ([#56550](https://github.com/kubernetes/kubernetes/pull/56550), [@andyzhangx](https://github.com/andyzhangx))
* Retry 'connection refused' errors when setting up clusters on GCE. ([#57394](https://github.com/kubernetes/kubernetes/pull/57394), [@mborsz](https://github.com/mborsz))

$(TARGET_OUT)/istio_is_init: bin/init.sh istio.deps | $(TARGET_OUT)
	@# Add a retry, as occasionally we see transient connection failures to GCS
	@# Like `curl: (56) OpenSSL SSL_read: SSL_ERROR_SYSCALL, errno 104`
	TARGET_OUT=$(TARGET_OUT) ISTIO_BIN=$(ISTIO_BIN) GOOS_LOCAL=$(GOOS_LOCAL) bin/retry.sh SSL_ERROR_SYSCALL bin/init.sh
	touch $(TARGET_OUT)/istio_is_init

.PHONY: init-ztunnel-rs
init-ztunnel-rs:

        .body("after misdirect")
        .build(),
    )

    // Seed the connection pool.
    assert200Http2Response(execute(url), server.hostName)

    // Use the coalesced connection which should retry on a fresh connection.
    val sanUrl =
      url.newBuilder()
        .host("san.com")
        .build()
    execute(sanUrl).use { response ->
      assertThat(response.code).isEqualTo(200)

- Kubeadm: apply further improvements to the tentative support for concurrent etcd member join. Fixes a bug where multiple members can receive the same hostname. Increase the etcd client dial timeout and retry timeout for add/remove/... operations. ([#87505](https://github.com/kubernetes/kubernetes/pull/87505), [@neolit123](https://github.com/neolit123)) [SIG Cluster Lifecycle]

## Changelog since v1.8.5

### Other notable changes

* change default azure file/dir mode to 0755 ([#56551](https://github.com/kubernetes/kubernetes/pull/56551), [@andyzhangx](https://github.com/andyzhangx))
* Retry 'connection refused' errors when setting up clusters on GCE. ([#57394](https://github.com/kubernetes/kubernetes/pull/57394), [@mborsz](https://github.com/mborsz))

### Bug or Regression

- An issue preventing the daemonset controller to retry on status updates that failed because of conflicts is now fixed. ([#107639](https://github.com/kubernetes/kubernetes/pull/107639), [@ialidzhikov](https://github.com/ialidzhikov)) [SIG Apps]

* vSphere cloud provider: Fix volume detach on node failure. ([#45569](https://github.com/kubernetes/kubernetes/pull/45569), [@divyenpatel](https://github.com/divyenpatel))
* Job controller: Retry with rate-limiting if a Pod create/delete operation fails. ([#45870](https://github.com/kubernetes/kubernetes/pull/45870), [@tacy](https://github.com/tacy))