当我更新 kubernetes(
v1.28.3│ Warning FailedCreatePodSandBox 10m kubelet Failed to create pod sandbox: rpc error: code = Unknown desc = failed to get sandbox image "registry.k8s.io/pause:3.6": fail │
│ ed to pull image "registry.k8s.io/pause:3.6": failed to pull and unpack image "registry.k8s.io/pause:3.6": failed to resolve reference "registry.k8s.io/pause:3.6": failed to do request: Head "https://u │
│ s-west2-docker.pkg.dev/v2/k8s-artifacts-prod/images/pause/manifests/3.6": dial tcp 142.251.8.82:443: i/o timeout │
│ Warning FailedCreatePodSandBox 6m20s kubelet Failed to create pod sandbox: rpc error: code = DeadlineExceeded desc = failed to get sandbox image "registry.k8s.io/pause:3 │
│ .6": failed to pull image "registry.k8s.io/pause:3.6": failed to pull and unpack image "registry.k8s.io/pause:3.6": failed to resolve reference "registry.k8s.io/pause:3.6": failed to do request: Head " │
│ https://us-west2-docker.pkg.dev/v2/k8s-artifacts-prod/images/pause/manifests/3.6": dial tcp 173.194.174.82:443: i/o timeout │
│ Warning FailedCreatePodSandBox 3m23s kubelet Failed to create pod sandbox: rpc error: code = Unknown desc = failed to get sandbox image "registry.k8s.io/pause:3.6": fail │
│ ed to pull image "registry.k8s.io/pause:3.6": failed to pull and unpack image "registry.k8s.io/pause:3.6": failed to resolve reference "registry.k8s.io/pause:3.6": failed to do request: Head "https://u │
│ s-west2-docker.pkg.dev/v2/k8s-artifacts-prod/images/pause/manifests/3.6": dial tcp 173.194.174.82:443: i/o timeout │
│ Warning FailedCreatePodSandBox 2m38s (x4 over 8m30s) kubelet Failed to create pod sandbox: rpc error: code = DeadlineExceeded desc = failed to get sandbox image "registry.k8s.io/pause:3 │
│ .6": failed to pull image "registry.k8s.io/pause:3.6": failed to pull and unpack image "registry.k8s.io/pause:3.6": failed to resolve reference "registry.k8s.io/pause:3.6": failed to do request: Head " │
│ https://us-west2-docker.pkg.dev/v2/k8s-artifacts-prod/images/pause/manifests/3.6": dial tcp 142.251.8.82:443: i/o timeout │
│ Warning FailedCreatePodSandBox 115s kubelet Failed to create pod sandbox: rpc error: code = Unknown desc = failed to get sandbox image "registry.k8s.io/pause:3.6": fail │
│ ed to pull image "registry.k8s.io/pause:3.6": failed to pull and unpack image "registry.k8s.io/pause:3.6": failed to resolve reference "registry.k8s.io/pause:3.6": failed to do request: Head "https://u │
│ s-west2-docker.pkg.dev/v2/k8s-artifacts-prod/images/pause/manifests/3.6": dial tcp 64.233.188.82:443: i/o timeout │
│ Warning FailedCreatePodSandBox 28s (x9 over 12m) kubelet Failed to create pod sandbox: rpc error: code = DeadlineExceeded desc = failed to get sandbox image "registry.k8s.io/pause:3 │
│ .6": failed to pull image "registry.k8s.io/pause:3.6": failed to pull and unpack image "registry.k8s.io/pause:3.6": failed to resolve reference "registry.k8s.io/pause:3.6": failed to do request: Head " │
│ https://us-west2-docker.pkg.dev/v2/k8s-artifacts-prod/images/pause/manifests/3.6": dial tcp 64.233.188.82:443: i/o timeout
我已经尝试过像这样拉取图像：
ctr -n=k8s.io image pull k8s.m.daocloud.io/pause:3.6
ctr -n=k8s.io images tag k8s.m.daocloud.io/pause:3.6 registry.k8s.io/pause:3.6
这种方式工作了一段时间，下次更新pod时，又会出现这个错误。我应该怎么做才能永久解决这个问题？
看起来错误的根本原因是/var分区可能有足够的可用空间（用df -h检查），但是在升级过程中，一旦加载新版本映像但在删除旧版本映像之前，它的使用率可能超过 80%。这是问题的主要根源。
因此，Kubernetes 可能会“清理” 或 “垃圾收集” 识别为未使用的图像，包括必要但不活动的 “暂停” 图像。
根据此 Medium 博客，作者：Richard Durso
要释放磁盘空间，请使用命令 crictl images —prune，有时它可能不会清除正在使用的所有内容。它甚至清除了 Kubernetes 暂停容器镜像。
kubelet 标志定义为：
image-gc-high-threshold：触发图像垃圾收集的磁盘使用百分比。默认为 85%。
image-gc-low-threshold：图像垃圾收集尝试释放的磁盘使用百分比。默认为 80%。
配置 Kubernetes 垃圾收集应该是维护磁盘空间的主要方式。但是，您仍然偶尔使用 CRI Purge 脚本来检查缓存了哪些映像，并在需要时手动执行它，例如在应用 Ubuntu 补丁之前进行节点维护时。
您可以尝试将 ImagePullPolicy 设置为 IfNotPresent (
spec.containers.imagePullPolicy: "ifNotPresent