Reduce CPU pinning fragmentation #384
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
liranr23
requested review from
ahadas,
bennyz,
emesika,
mwperina,
michalskrivanek,
oliel,
sgratch and
didib
as code owners
|
A bit note: |
ljelinkova
reviewed
May 23, 2022
There was a problem hiding this comment.
Usually, I am a bit skeptical about the usefulness of the unit tests but in case of the VmsCpuPinningPolicyComparator I think it might be actually useful to provide some VMs and see what is the expected output for different pinning, pinning policies and topologies.
...org/ovirt/engine/core/common/businessentities/comparators/VmsCpuPinningPolicyComparator.java
Outdated
Show resolved
Hide resolved
...org/ovirt/engine/core/common/businessentities/comparators/VmsCpuPinningPolicyComparator.java
Outdated
Show resolved
Hide resolved
...anager/modules/bll/src/main/java/org/ovirt/engine/core/bll/scheduling/SchedulingManager.java
Outdated
Show resolved
Hide resolved
yeah, in this case it will also serve as a documentation |
ahadas
reviewed
May 23, 2022
liranr23
commented
May 24, 2022
...anager/modules/bll/src/main/java/org/ovirt/engine/core/bll/scheduling/SchedulingManager.java
Outdated
Show resolved
Hide resolved
...org/ovirt/engine/core/common/businessentities/comparators/VmsCpuPinningPolicyComparator.java
Outdated
Show resolved
Hide resolved
...anager/modules/bll/src/main/java/org/ovirt/engine/core/bll/scheduling/SchedulingManager.java
Outdated
Show resolved
Hide resolved
...org/ovirt/engine/core/common/businessentities/comparators/VmsCpuPinningPolicyComparator.java
Outdated
Show resolved
Hide resolved
...org/ovirt/engine/core/common/businessentities/comparators/VmsCpuPinningPolicyComparator.java
Outdated
Show resolved
Hide resolved
|
Based on the offline discussion:
|
liranr23
force-pushed
the
order_schedule_vms
branch
from
d937243
to
196bcfe
Compare
|
Need to verify, and add unit tests. But this is how it looks now. |
liranr23
force-pushed
the
order_schedule_vms
branch
2 times, most recently
from
a03ea57
to
dbcb0cd
Compare
liranr23
commented
Jun 6, 2022
...org/ovirt/engine/core/common/businessentities/comparators/VmsCpuPinningPolicyComparator.java
Outdated
Show resolved
Hide resolved
ahadas
reviewed
Jun 6, 2022
...anager/modules/bll/src/main/java/org/ovirt/engine/core/bll/scheduling/SchedulingManager.java
Outdated
Show resolved
Hide resolved
...anager/modules/bll/src/main/java/org/ovirt/engine/core/bll/scheduling/SchedulingManager.java
Outdated
Show resolved
Hide resolved
...org/ovirt/engine/core/common/businessentities/comparators/VmsCpuPinningPolicyComparator.java
Outdated
Show resolved
Hide resolved
liranr23
force-pushed
the
order_schedule_vms
branch
2 times, most recently
from
1ea8bcd
to
2076b6a
Compare
ljelinkova
reviewed
Jun 7, 2022
...les/common/src/main/java/org/ovirt/engine/core/common/businessentities/CpuPinningPolicy.java
Show resolved
Hide resolved
ljelinkova
approved these changes
Jun 7, 2022
This patch sorts the VM list on schedule phase before doing the dedicated CPU pinning. This change tries to reduce the fragmentation that may be caused when allocating a smaller VM in terms of the policy and CPU topology (a less restricted) when having a bigger one. Scheduling the smaller VM first may cause less physical resources needed by the larger VM because we try to enhance performance in the allocation logic. Now, the VMs will be sorted based on the policy first and if needed by the CPU topology, letting the more restricted VM to be scheduled first. Change-Id: Ia3fc28511c147d43da9cf8d86d50ca17f1e3f0e1 Signed-off-by: Liran Rotenberg <lrotenbe@redhat.com>
liranr23
force-pushed
the
order_schedule_vms
branch
from
2076b6a
to
4cc058f
Compare
|
/ost |
|
OST failed on UI test, unrelated to the patch. |
ahadas
approved these changes
Jun 9, 2022
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This patch sorts the VM list on schedule phase before doing the
dedicated CPU pinning.
This change tries to reduce the fragmentation that may be caused when
allocating a smaller VM in terms of the policy and CPU topology (a less
restricted) when having a bigger one. Scheduling the smaller VM first
may cause less physical resources needed by the larger VM because we try
to enhance performance in the allocation logic.
Now, the VMs will be sorted based on the policy first and if needed by
the CPU topology, letting the more restricted VM to be scheduled first.
Change-Id: Ia3fc28511c147d43da9cf8d86d50ca17f1e3f0e1
Signed-off-by: Liran Rotenberg lrotenbe@redhat.com