|
本文档采用自动化机器翻译技术翻译。 尽管我们力求提供准确的译文，但不对翻译内容的完整性、准确性或可靠性作出任何保证。 若出现任何内容不一致情况，请以原始 英文 版本为准，且原始英文版本为权威文本。 |
故障排除重置
每个 MachineInventory 带有 elemental.cattle.io/resettable: "true" 注释的项在删除时将触发重置计划的执行。
只有当 machineinventory.elemental.cattle.io 运行在机器上的 elemental-system-agent 成功执行该计划后，清理器才会被移除。
您可以通过检查 MachineInventory 来调查其为何尚未被删除：
apiVersion: elemental.cattle.io/v1beta1
kind: MachineInventory
metadata:
# deletionTimestamp has been set. This object has been marked for deletion.
deletionTimestamp: "2023-08-04T08:41:25Z"
annotations:
# `resettable` annotation is enabled.
# This means the machine has to go through reset, before deletion of this object.
elemental.cattle.io/resettable: "true"
# `machineinventory.elemental.cattle.io` finalizer is set.
# The `elemental-operator` is going to create a reset plan for the machine to execute.
# After successful execution of the reset plan, the finalizer is removed and the object will be deleted.
finalizers:
- machineinventory.elemental.cattle.io
status:
conditions:
# Most recent condition shows that the MachineInventory is waiting for a plan to be applied.
- lastTransitionTime: "2023-08-04T08:41:25Z"
message: waiting for plan to be applied
reason: WaitingForPlan
status: "False"
type: Ready
# The plan to be executed is referenced.
# Normally it has the same name of the MachineInventory and lives within the same namespace.
plan:
checksum: 5aba8b6b3161bc52d8953b2428e54ecda3b59e8e0043b49d761d1e79174eded6
secretRef:
name: m-bf1008a1-61d6-4355-b5f5-f7d1c527affe
namespace: fleet-default您还可以检查引用的计划 Secret。
请注意，在机器上运行的 elemental-system-agent 正在监视此秘密，并应执行该计划。
您还可以从机器日志 journalctl -u elemental-system-agent -f 监控其进度。
apiVersion: v1
kind: Secret
# This is a `elemental.cattle.io/plan` secret plan.
type: elemental.cattle.io/plan
metadata:
annotations:
# This is a `reset` plan type.
elemental.cattle.io/plan.type: reset
labels:
elemental.cattle.io/managed: "true"
name: m-bf1008a1-61d6-4355-b5f5-f7d1c527affe
namespace: fleet-default
# It is owned by the `MachineInventory` waiting for deletion.
ownerReferences:
- apiVersion: elemental.cattle.io/v1beta1
controller: true
kind: MachineInventory
name: m-bf1008a1-61d6-4355-b5f5-f7d1c527affe
uid: 5aa3863c-63a5-4cb9-91fd-7a45191d4842
data:
# The plan has not been applied yet.
applied-checksum: ""
# It also hasn't failed.
failed-checksum: ""
# The actual plan to be executed, base64 encoded.
plan: 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由 elemental-operator 创建的计划应包含以下指令：
{
"files": [
// A cloud-init config file is created on the default /oem directory.
// This config will be executed once in recovery mode.
{
"content": "bmFtZTogRWxlbWVudGFsIFJlc2V0CnN0YWdlczoKICAgIG5ldHdvcmsuYWZ0ZXI6CiAgICAgICAgLSBjb21tYW5kczoKICAgICAgICAgICAgLSBlbGVtZW50YWwtcmVnaXN0ZXIgLS1kZWJ1ZyAtLXJlc2V0CiAgICAgICAgICBpZjogJ1sgLWYgL3J1bi9jb3MvcmVjb3ZlcnlfbW9kZSBdJwogICAgICAgICAgbmFtZTogUnVucyBlbGVtZW50YWwgcmVzZXQK",
"path": "/oem/reset-cloud-config.yaml",
"permissions": "0600"
}
],
"instructions": [
{
"name": "configure next boot to recovery mode",
"args": [
"/oem/grubenv",
"set",
"next_entry=recovery"
],
"command": "grub2-editenv"
},
{
"name": "schedule reboot",
"args": [
"-r",
"+1"
],
"command": "shutdown"
}
]
}如果 elemental-system-agent 成功执行了该计划，machineinventory.elemental.cattle.io 清理器将在 MachineInventory 上被移除，MachineInventory 将被删除。
请注意，这并不表示机器已经完全重置。
这是当前实现的一个限制，最终会有所改善，以便能够完全跟踪重置状态。
然而，在这个阶段，我们确实期望主机进行重启，并以恢复模式重启。
一旦进入恢复模式，cos-setup-network 应执行已写入 /oem/reset-cloud-config.yaml 的 cloud-init 配置。
您可以使用 journalctl -u cos-setup-network -f 监控状态。
cloud-init 指令应如下所示：
name: {elemental-product-name} Reset
stages:
network.after:
- if: '[ -f /run/cos/recovery_mode ]'
name: Runs elemental reset
commands:
- systemctl start elemental-register-resetelemental-register CLI 将作为新机器向 elemental-operator 注册。这将导致创建一个新的 MachineInventory 对象。
远程 MachineRegistration 配置也将被获取以应用重置选项，例如 reset-persistent、reset-oem 或电源设置，reboot 或 power-off。
重置后，根据设置，机器应关闭或重启，并准备在新集群中被采用。