Document URL: Section Number and Name: Describe the issue: terminationGracePeriodSeconds value is set to 0 (force stop) in common templates. If a VM is created using common templates, or from the UI using the wizard, terminationGracePeriodSeconds should be updated before the VM is started. Suggestions for improvement: Additional information:
@rmohr - can you lease suggest a recommended value?
That depends on the operating system. For many linux based systems the default of 30 seconds is good enough. For some windows versions which tend to do extensive updates when you reboot the OS, hours could be a good choice.
Requesting the following information @rmohr: The default value in the common templates is zero in common templates? We recommend that the value be modified and set to 30 seconds for all Linux operating systems, true? For Windows environments with extensive updates etc, we recommend the value be modified to how many hours?
(In reply to Lavanya Mandavilli from comment #3) > Requesting the following information @rmohr: > > The default value in the common templates is zero in common templates? > > We recommend that the value be modified and set to 30 seconds for all Linux > operating systems, true? It really depends on what you do. In many cases it is enough already. > > For Windows environments with extensive updates etc, we recommend the value > be modified to how many hours? Also here, depends on what the OS does, but it can take a long time until a windows machine shuts down voluntarily. I don't have exact values at hand to recommend since it is really very specific. I would really make with all suggestions clear that this is not something which we can define 100% correct up front.
I created a PR. Adding the link shortly.
Known Issue PR: https://github.com/openshift/openshift-docs/pull/24253 (Tagged Roman for review on PR) Should we also add an admonition to the docs until the fix is in?
This was verified by Nelly as part of a pre-release sweep in https://issues.redhat.com/browse/CNV-4934 The release note can be viewed here: https://docs.openshift.com/container-platform/4.5/virt/virt-2-4-release-notes.html#virt-2-4-known-issues @Ruth: Does this also require a admonition in the docs, or will this be fixed in the templates soon (and known issue will suffice)?
@Andrew: As this may have severe impact on VMs, I think if should include one.
Thanks Ruth. I've added an Important admonition at the end of the procedure for creating a VM using the VM wizard. it links to the workaround in the release notes. PR: https://github.com/openshift/openshift-docs/pull/24527 (There's no preview build but I've added screenshot of a local build to the PR description) Our content for creating a VM in the CLI is about posting a VM manifest to the cluster. We do not mention common-templates in the CLI pathway so no admonition added there. Please note, this is only for the enterprise-4.5 docs as it is expected the fix will be merged before 2.4.1
Reviewed the patch, LGTM. Moving to "verified".
The docs PR was merged and is now available: https://docs.openshift.com/container-platform/4.5/virt/virtual_machines/virt-create-vms.html