github/kata-containers
1
0
Fork 2
mirror of https://github.com/kata-containers/kata-containers.git synced 2025-09-17 14:58:16 +00:00
Code Issues Projects Releases Wiki Activity

runtime: Improve vCPU allocation for the VMMs

Browse Source 
First of all, this is a controversial piece, and I know that.

In this commit we're trying to make a less greedy approach regards the
amount of vCPUs we allocate for the VMM, which will be advantageous
mainly when using the `static_sandbox_resource_mgmt` feature, which is
used by the confidential guests.

The current approach we have basically does:
* Gets the amount of vCPUs set in the config (an integer)
* Gets the amount of vCPUs set as limit (an integer)
* Sum those up
* Starts / Updates the VMM to use that total amount of vCPUs

The fact we're dealing with integers is logical, as we cannot request
500m vCPUs to the VMMs.  However, it leads us to, in several cases, be
wasting one vCPU.

Let's take the example that we know the VMM requires 500m vCPUs to be
running, and the workload sets 250m vCPUs as a resource limit.

In that case, we'd do:
* Gets the amount of vCPUs set in the config: 1
* Gets the amount of vCPUs set as limit: ceil(0.25)
* 1 + ceil(0.25) = 1 + 1 = 2 vCPUs
* Starts / Updates the VMM to use 2 vCPUs

With the logic changed here, what we're doing is considering everything
as float till just before we start / update the VMM. So, the flow
describe above would be:
* Gets the amount of vCPUs set in the config: 0.5
* Gets the amount of vCPUs set as limit: 0.25
* ceil(0.5 + 0.25) = 1 vCPUs
* Starts / Updates the VMM to use 1 vCPUs

In the way I've written this patch we introduce zero regressions, as
the default values set are still the same, and those will only be
changed for the TEE use cases (although I can see firecracker, or any
other user of `static_sandbox_resource_mgmt=true` taking advantage of
this).

There's, though, an implicit assumption in this patch that we'd need to
make explicit, and that's that the default_vcpus / default_memory is the
amount of vcpus / memory required by the VMM, and absolutely nothing
else.  Also, the amount set there should be reflected in the
podOverhead for the specific runtime class.

One other possible approach, which I am not that much in favour of
taking as I think it's **less clear**, is that we could actually get the
podOverhead amount, subtract it from the default_vcpus (treating the
result as a float), then sum up what the user set as limit (as a float),
and finally ceil the result.  It could work, but IMHO this is **less
clear**, and **less explicit** on what we're actually doing, and how the
default_vcpus / default_memory should be used.

Fixes: #6909

Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Signed-off-by: Christophe de Dinechin <dinechin@redhat.com>
This commit is contained in:
Fabiano Fidêncio
2023-06-30 12:17:51 +00:00
parent b0157ad73a
commit e477ed0e86
29 changed files with 120 additions and 102 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
src/runtime/virtcontainers/qemu.go
View File

@@ -315,7 +315,7 @@ func (q *qemu) setup(ctx context.Context, id string, hypervisorConfig *Hyperviso
}
func (q *qemu) cpuTopology() govmmQemu.SMP {
return q.arch.cpuTopology(q.config.NumVCPUs, q.config.DefaultMaxVCPUs)
return q.arch.cpuTopology(q.config.NumVCPUs(), q.config.DefaultMaxVCPUs)
}
func (q *qemu) memoryTopology() (govmmQemu.Memory, error) {
@@ -630,7 +630,7 @@ func (q *qemu) CreateVM(ctx context.Context, id string, network Network, hypervi
// we must ensure we use something like:
// `...,sockets=1,cores=numvcpus,threads=1,...`
smp.Sockets = 1
smp.Cores = q.config.NumVCPUs
smp.Cores = q.config.NumVCPUs()
smp.Threads = 1
}
}
@@ -1590,7 +1590,7 @@ func (q *qemu) hotplugAddBlockDevice(ctx context.Context, drive *config.BlockDri
return err
}
queues := int(q.config.NumVCPUs)
queues := int(q.config.NumVCPUs())
if err = q.qmpMonitorCh.qmp.ExecutePCIDeviceAdd(q.qmpMonitorCh.ctx, drive.ID, devID, driver, addr, bridge.ID, romFile, queues, true, defaultDisableModern); err != nil {
return err
@@ -1922,7 +1922,7 @@ func (q *qemu) hotplugNetDevice(ctx context.Context, endpoint Endpoint, op Opera
addr := "00"
bridgeID := fmt.Sprintf("%s%d", config.PCIeRootPortPrefix, len(config.PCIeDevices[config.RootPort]))
config.PCIeDevices[config.RootPort][devID] = true
return q.qmpMonitorCh.qmp.ExecuteNetPCIDeviceAdd(q.qmpMonitorCh.ctx, tap.Name, devID, endpoint.HardwareAddr(), addr, bridgeID, romFile, int(q.config.NumVCPUs), defaultDisableModern)
return q.qmpMonitorCh.qmp.ExecuteNetPCIDeviceAdd(q.qmpMonitorCh.ctx, tap.Name, devID, endpoint.HardwareAddr(), addr, bridgeID, romFile, int(q.config.NumVCPUs()), defaultDisableModern)
}
addr, bridge, err := q.arch.addDeviceToBridge(ctx, tap.ID, types.PCI)
@@ -1954,9 +1954,9 @@ func (q *qemu) hotplugNetDevice(ctx context.Context, endpoint Endpoint, op Opera
}
if machine.Type == QemuCCWVirtio {
devNoHotplug := fmt.Sprintf("fe.%x.%x", bridge.Addr, addr)
return q.qmpMonitorCh.qmp.ExecuteNetCCWDeviceAdd(q.qmpMonitorCh.ctx, tap.Name, devID, endpoint.HardwareAddr(), devNoHotplug, int(q.config.NumVCPUs))
return q.qmpMonitorCh.qmp.ExecuteNetCCWDeviceAdd(q.qmpMonitorCh.ctx, tap.Name, devID, endpoint.HardwareAddr(), devNoHotplug, int(q.config.NumVCPUs()))
}
return q.qmpMonitorCh.qmp.ExecuteNetPCIDeviceAdd(q.qmpMonitorCh.ctx, tap.Name, devID, endpoint.HardwareAddr(), addr, bridge.ID, romFile, int(q.config.NumVCPUs), defaultDisableModern)
return q.qmpMonitorCh.qmp.ExecuteNetPCIDeviceAdd(q.qmpMonitorCh.ctx, tap.Name, devID, endpoint.HardwareAddr(), addr, bridge.ID, romFile, int(q.config.NumVCPUs()), defaultDisableModern)
}
if err := q.arch.removeDeviceFromBridge(tap.ID); err != nil {
@@ -2687,9 +2687,9 @@ func calcHotplugMemMiBSize(mem uint32, memorySectionSizeMB uint32) (uint32, erro
}
func (q *qemu) ResizeVCPUs(ctx context.Context, reqVCPUs uint32) (currentVCPUs uint32, newVCPUs uint32, err error) {
currentVCPUs = q.config.NumVCPUs + uint32(len(q.state.HotpluggedVCPUs))
currentVCPUs = q.config.NumVCPUs() + uint32(len(q.state.HotpluggedVCPUs))
newVCPUs = currentVCPUs
switch {
case currentVCPUs < reqVCPUs:
//hotplug
@@ -2716,6 +2716,7 @@ func (q *qemu) ResizeVCPUs(ctx context.Context, reqVCPUs uint32) (currentVCPUs u
}
newVCPUs -= vCPUsRemoved
}
return currentVCPUs, newVCPUs, nil
}