diff options
| author | Adin Scannell <ascannell@google.com> | 2018-04-28 18:05:12 -0700 |
|---|---|---|
| committer | Shentubot <shentubot@google.com> | 2018-04-28 18:05:59 -0700 |
| commit | 913aa0a24dc3ed6cb71971e56c6ae363e8d0113d (patch) | |
| tree | adebe9adf85cf170e8a8b2463d5339629c87a5ab | |
| parent | 54a20025b15795bf3d6bd2435b8e83dbbca55a64 (diff) | |
Restore markdown images that work universally
PiperOrigin-RevId: 194676199
Change-Id: Ibb9257a5504b10c4469a57ba27cd866f2d660fd8
| -rw-r--r-- | README.md | 14 |
1 files changed, 7 insertions, 7 deletions
@@ -45,7 +45,7 @@ performance (though nested virtualization may bring challenges in this area), but for containers it often requires additional proxies and agents, and may require a larger resource footprint and slower start-up times. -<p align="center"><img src="g3doc/Machine-Virtualization.png"></p> + **Rule-based execution**, such as [seccomp][seccomp], [SELinux][selinux] and [AppArmor][apparmor], allows the specification of a fine-grained security policy @@ -57,14 +57,16 @@ However, in practice it can be extremely difficult (if not impossible) to reliably define a policy for arbitrary, previously unknown applications, making this approach challenging to apply universally. + + Rule-based execution is often combined with additional layers for defense-in-depth. -<p align="center"><img src="g3doc/Rule-Based-Execution.png"></p> - **gVisor** provides a third isolation mechanism, distinct from those mentioned above. + + gVisor intercepts application system calls and acts as the guest kernel, without the need for translation through virtualized hardware. gVisor may be thought of as either a merged guest kernel and VMM, or as seccomp on steroids. This @@ -76,8 +78,6 @@ reduced application compatibility and higher per-system call overhead. On top of this, gVisor employs rule-based execution to provide defense-in-depth (details below). -<p align="center"><img src="g3doc/Layers.png"></p> - gVisor's approach is similar to [User Mode Linux (UML)][uml], although UML virtualizes hardware internally and thus provides a fixed resource footprint. @@ -106,6 +106,8 @@ application to directly control the system calls it makes. ### File System Access + + In order to provide defense-in-depth and limit the host system surface, the gVisor container runtime is normally split into two separate processes. First, the *Sentry* process includes the kernel and is responsible for executing user @@ -119,8 +121,6 @@ access itself. Furthermore, the Sentry runs in an empty user namespace, and the system calls made by gVisor to the host are restricted using seccomp filters in order to provide defense-in-depth. -<p align="center"><img src="g3doc/Sentry-Gofer.png"></p> - ### Network Access The Sentry implements its own network stack (also written in Go) called |