QEMU: enhance documentation and check for /dev/kvm #374
Conversation
| section in the Clear Linux documentation contains further information | ||
| about enabling KVM. | ||
| KVM must be enabled. Usually this is the case when `/dev/kvm` exists. | ||
| The current user does not need the privileges to use KVM and QEMU |
There was a problem hiding this comment.
Nitpick question, do we need to decide to use does not or doesn't and stick to it trough the md documents? or does not matter that much?
It came to my attention because in this line does not is used and the next one is doesn't
(And I think this will apply for do not and don't too and all the shorter version of ENG words)
|
Marcela Morales Q <notifications@github.com> writes:
marcemq commented on this pull request.
> @@ -783,11 +783,14 @@ available.
This is known to work on a Linux development host system. The user
must be allowed to use Docker.
-KVM must be enabled and the user must be allowed to use it. Usually this
-is done by adding the user to the `kvm` group. The
-["Install QEMU-KVM"](https://clearlinux.org/documentation/clear-linux/get-started/virtual-machine-install/kvm)
-section in the Clear Linux documentation contains further information
-about enabling KVM.
+KVM must be enabled. Usually this is the case when `/dev/kvm` exists.
+The current user does not need the privileges to use KVM and QEMU
Nitpick question, do we need to decide to use **does not** or
**doesn't** and stick to it trough the `md` documents? or does not
matter that much?
I learned that "doesn't" and "don't" are colloquial and not to be used
in articles or scientific papers. But it seems that documentation
nowadays leans more towards colloquial than scientific.
I think it doesn't matter that much in the end. In this case, spelling
it out has the additional advantage that it emphasizes the "not", as in
"does *not* need the privileges".
|
pohly
changed the title
|
Don't merge, the |
pohly
force-pushed
the
kvm-error-handline
branch
from
ad3062d to
b1dc655
Compare
pohly
force-pushed
the
kvm-error-handline
branch
from
b1dc655 to
6230570
Compare
|
I'm still not sure why the I also added a commit which enables nested virtualization. It's a small additional change that touches the same documentation section; I didn't want to create another conflicting PR for that. |
pohly
force-pushed
the
kvm-error-handline
branch
from
6230570 to
9a63ca2
Compare
pohly
changed the title
|
@avalluri please also check the second commit. |
|
@avalluri: ping? We currently have clean test results. |
Not having KVM enabled in a GCE machine is a real issue that a user ran into. Error handling in this case is poor, GoVM doesn't print any error and all that the user gets is an error about timing out waiting for ssh. The updated README.md explains the real requirements better and warns about this pitfall. The script itself checks very early for /dev/kvm and bails out with an error message when it is missing.
pohly
force-pushed
the
kvm-error-handline
branch
from
9a63ca2 to
2a9b234
Compare
It was broken on "devel" by an an earlier PR (#473) and has already been fixed there by #481. This change here is fine, it was just based on that broken revision of "devel". I rebased, so now "view file" shows a clean version. |
README.md
Outdated
| this time, Kata Containers up to and including 1.9.1 is [not | ||
| compatible with | ||
| PMEM-CSI](https://github.com/intel/pmem-csi/issues/303) because | ||
| volumes are not passed in as PMEM, but it [can be |
There was a problem hiding this comment.
but it can be installed...
I am not clear what exactly 'it' refers to in this sentence? Can you please check if any rephrase required or just my in-ablity to understand :-}
There was a problem hiding this comment.
it = kata containers. I'll replace the "it".
Nested virtualization is needed for Kata Containers. This works on hosts which have kvm_intel loaded with nested=1. Might also need a recent enough host CPU.
pohly
force-pushed
the
kvm-error-handline
branch
from
2a9b234 to
7f0f6ab
Compare
Not having KVM enabled in a GCE machine is a real issue that a user
ran into. Error handling in this case is poor, GoVM doesn't print any
error and all that the user gets is an error about timing out waiting
for ssh.
The updated README.md explains the real requirements better and warns
about this pitfall. The script itself checks very early for /dev/kvm
and bails out with an error message when it is missing.