@@ -281,42 +281,35 @@ Memory limit superior to request can expose the whole node to OOM issues.
281
281
<!--
282
282
### Enabling Seccomp
283
283
284
- Seccomp can improve the security of your workloads by reducing the Linux kernel
285
- syscall attack surface available inside containers. The seccomp filter mode
286
- leverages BPF to create an allow or deny list of specific syscalls, named
287
- profiles. Those seccomp profiles can be enabled on individual workloads,
288
- [a security tutorial is available](/docs/tutorials/security/seccomp/). In
289
- addition, the [Kubernetes Security Profiles Operator](https://github.com/kubernetes-sigs/security-profiles-operator)
290
- is a project to facilitate the management and use of seccomp in clusters.
284
+ Seccomp stands for secure computing mode and has been a feature of the Linux kernel since version 2.6.12.
285
+ It can be used to sandbox the privileges of a process, restricting the calls it is able to make
286
+ from userspace into the kernel. Kubernetes lets you automatically apply seccomp profiles loaded onto
287
+ a node to your Pods and containers.
288
+
289
+ Seccomp can improve the security of your workloads by reducing the Linux kernel syscall attack
290
+ surface available inside containers. The seccomp filter mode leverages BPF to create an allow or
291
+ deny list of specific syscalls, named profiles.
292
+
293
+ Since Kubernetes 1.27, you can enable the use of `RuntimeDefault` as the default seccomp profile
294
+ for all workloads. A [security tutorial](/docs/tutorials/security/seccomp/) is available on this
295
+ topic. In addition, the
296
+ [Kubernetes Security Profiles Operator](https://github.com/kubernetes-sigs/security-profiles-operator)
297
+ is a project that facilitates the management and use of seccomp in clusters.
291
298
-->
292
299
### 启用 Seccomp {#enabling-seccomp}
293
300
294
- <!-- 按照英文原文翻译比较啰嗦,本小段是英文原文结合 Seccomp 简洁翻译的 -->
301
+ Seccomp 代表安全计算模式(Secure computing mode),这是一个自 Linux 内核版本 2.6.12 被加入的特性。
302
+ 它可以将进程的特权沙箱化,来限制从用户空间发起的对内核的调用。
303
+ Kubernetes 允许你将加载到节点上的 Seccomp 配置文件自动应用于你的 Pod 和容器。
304
+
295
305
Seccomp 通过减少容器内对 Linux 内核的系统调用(System Call)以缩小攻击面,从而提高工作负载的安全性。
296
- Seccomp 过滤器模式借助 BPF 创建了配置文件(Profile),文件中设置对具体系统调用的允许或拒绝,
297
- 可以针对单个工作负载上启用这类 Seccomp 配置文件。你可以阅读相应的[ 安全教程] ( /zh-cn/docs/tutorials/security/seccomp/ ) 。
306
+ Seccomp 过滤器模式借助 BPF 创建具体系统调用的允许清单或拒绝清单,名为配置文件(Profile)。
307
+
308
+ 从 Kubernetes 1.27 开始,你可以将 ` RuntimeDefault ` 设置为工作负载的默认 Seccomp 配置。
309
+ 你可以阅读相应的[ 安全教程] ( /zh-cn/docs/tutorials/security/seccomp/ ) 。
298
310
此外,[ Kubernetes Security Profiles Operator] ( https://github.com/kubernetes-sigs/security-profiles-operator )
299
311
是一个方便在集群中管理和使用 Seccomp 的项目。
300
312
301
- <!--
302
- For historical context, please note that Docker has been using
303
- [a default seccomp profile](https://docs.docker.com/engine/security/seccomp/)
304
- to only allow a restricted set of syscalls since 2016 from
305
- [Docker Engine 1.10](https://www.docker.com/blog/docker-engine-1-10-security/),
306
- but Kubernetes is still not confining workloads by default. The default seccomp
307
- profile can be found [in containerd](https://github.com/containerd/containerd/blob/main/contrib/seccomp/seccomp_default.go)
308
- as well. Fortunately, [Seccomp Default](/blog/2021/08/25/seccomp-default/), a
309
- new alpha feature to use a default seccomp profile for all workloads can now be
310
- enabled and tested.
311
- -->
312
- 从历史背景看,请注意 Docker 自 2016 年以来一直使用[ 默认的 Seccomp 配置文件] ( https://docs.docker.com/engine/security/seccomp/ ) ,
313
- 仅允许来自 [ Docker Engine 1.10] ( https://www.docker.com/blog/docker-engine-1-10-security/ ) 的很小的一组系统调用,
314
- 但是,在默认情况下 Kubernetes 仍不限制工作负载。
315
- 默认的 Seccomp 配置文件也可以在
316
- [ containerd] ( https://github.com/containerd/containerd/blob/main/contrib/seccomp/seccomp_default.go ) 中找到。
317
- 幸运的是,[ Seccomp Default] ( /blog/2021/08/25/seccomp-default/ ) 可将默认的 Seccomp 配置文件用于所有工作负载,
318
- 这是一项新的 Alpha 功能,现在可以启用和测试了。
319
-
320
313
{{< note >}}
321
314
<!--
322
315
Seccomp is only available on Linux nodes.
0 commit comments