-
Notifications
You must be signed in to change notification settings - Fork 24
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
MTV-1211 Max concurrent virtual machine migrations #580
base: main
Are you sure you want to change the base?
Conversation
Your call as to what you incorporate. :-) I might change: "Maximum number of virtual machines (VMs) or disks per plan that can be migrated simultaneously." (---unless “can be” means that you may decide not to and have that option) "The maximum number of disks that can be transferred simultaneously." "The maximum number of VMs that can be migrated simultaneously." "…can be transferred simultaneously. In these migrations, the disks are migrated in parallel. This means that if the combined number of disks that you want to migrate is greater than the value of the setting, additional disks wait until the queue is free without regard for whether a VM has completely finished being migrated." "Once any of them has migrated, the 16th disk can be migrated, whether or not all the disks on VM A and VM B have finished migrating." (-- unless “can be migrated” means that it’s an option to migrate the disk) |
fd2de05
to
c9a5e37
Compare
@RichardHoch @mnecas |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM except the comments below.
But let's wait for @mnecas final approval as well.
I am a bit afraid that we would lose the previous settings during the upgrade. |
OK, so let's leave it as is for now. |
c9a5e37
to
2b03d58
Compare
@RichardHoch did you miss those 2 comments ? :) |
The 'controller_max_vm_inflight' parameter is related to simultaneous migration. |
LGTM |
LGTM |
Signed-off-by: RichardHoch <[email protected]>
fb2d81a
to
463292f
Compare
@duduvaa You replied LGTM to Andy's suggestion, but what about cold remote? Martin commented "For cold remote we use the CNV CDI so the max in flight references the number of disks" Does that fit your test results?
** Cold migration: *** To local {virt}: VMs for each ESXi host that can migrate simultaneously ** Warm migration: Disks for each ESXi host that can migrate simultaneously |
@RichardHoch , |
I tested cold migration to remote env. Richard, as you wrote *** To local {virt}: VMs for each ESXi host that can migrate simultaneously |
@RichardHoch I believe you asked about Redhat Virtualization/ Openstack not OCP-V |
@duduvaa Actually, I did mean OpenShift Virtualization, aka CNV, because it's a source provider, but we didn't define whether the parameter applies to disks or VMs when CNV is the source provider. |
|
In MTV, cnv cannot be source provider only destination
…On Mon, 6 Jan 2025, 13:10 David Vaanunu, ***@***.***> wrote:
@RichardHoch <https://github.com/RichardHoch> I believe you asked about
Redhat Virtualization/ Openstack not OCP-V
@duduvaa <https://github.com/duduvaa> Actually, I did mean OpenShift
Virtualization, aka CNV, because it's a source provider, but we didn't
define whether the parameter applies to disks or VMs when CNV is the source
provider.
@fabiand <https://github.com/fabiand> , @mnecas
<https://github.com/mnecas>
Could you relate to Richard's comment?
—
Reply to this email directly, view it on GitHub
<#580 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AHGJQIVDD2GAMNMH26C76LD2JJQCBAVCNFSM6AAAAABQ6X725GVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNZSHA4DQNZQGA>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
OCP can be both - src + dst. @mnecas mentioned that max inflight for CNV is about inflight VMs. |
We are using the |
@mnecas I have added the information for CNV as provider. Please review the PR again. Thanks. |
[id="max-concurrent-vms_{context}"] | ||
= Configuring the controller_max_vm_inflight label | ||
|
||
For {rhv-full} migrations, {osp} migrations, VMware warm migrations, or cold VMware migrations to a remote {ocp} cluster the setting `controller_max_vm_inflight` (shown in the UI as *Max concurrent virtual machine migrations*) specifies the maximum number of _disks_ that {project-first} can transfer simultaneously. In these migrations, {project-short} migrates the disks in parallel. This means that if the combined number of disks that you want to migrate is greater than the value of the setting, additional disks must wait until the queue is free, without regard for whether a VM has finished migrating. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I did more reading and found out that the rhv and osp also do it per VM.
I think the code is not correct and should be changed but this is the current status so I wanted to correct myself.
So the per disk applies only to remote vmware or warm vmware migrations.
@mnecas I made the changes you suggested. Please review this PR again. |
MTV 2.7
Resolves https://issues.redhat.com/browse/MTV-1211 by adding information about
controller_max_vm_inflight
(shown in the UI as Max concurrent virtual machine migrations).Previews: