Long running tests like TestDRA/all/DeviceBindingConditions (42.50s)
should run in parallel with other tests, otherwise the overall runtime is too
high.
This then must allow more pods per node to avoid blocking scheduling.
GatherAllocatedState and ListAllAllocatedDevices need to collect information
from different sources (allocated devices, in-flight claims), potentially even
multiple times (GatherAllocatedState first gets allocated devices, then the
capacities).
The underlying assumption that nothing bad happens in parallel is not always
true. The following log snippet shows how an update of the assume
cache (feeding the allocated devices tracker) and in-flight claims lands such
that GatherAllocatedState doesn't see the device in that claim as allocated:
dra_manager.go:263: I0115 15:11:04.407714 18778] scheduler: Starting GatherAllocatedState
...
allocateddevices.go:189: I0115 15:11:04.407945 18066] scheduler: Observed device allocation device="testdra-all-usesallresources-hvs5d.driver/worker-5/worker-5-device-094" claim="testdra-all-usesallresources-hvs5d/claim-0553"
dynamicresources.go:1150: I0115 15:11:04.407981 89109] scheduler: Claim stored in assume cache pod="testdra-all-usesallresources-hvs5d/my-pod-0553" claim="testdra-all-usesallresources-hvs5d/claim-0553" uid=<types.UID>: a84d3c4d-f752-4cfd-8993-f4ce58643685 resourceVersion="5680"
dra_manager.go:201: I0115 15:11:04.408008 89109] scheduler: Removed in-flight claim claim="testdra-all-usesallresources-hvs5d/claim-0553" uid=<types.UID>: a84d3c4d-f752-4cfd-8993-f4ce58643685 version="1211"
dynamicresources.go:1157: I0115 15:11:04.408044 89109] scheduler: Removed claim from in-flight claims pod="testdra-all-usesallresources-hvs5d/my-pod-0553" claim="testdra-all-usesallresources-hvs5d/claim-0553" uid=<types.UID>: a84d3c4d-f752-4cfd-8993-f4ce58643685 resourceVersion="5680" allocation=<
{
"devices": {
"results": [
{
"request": "req-1",
"driver": "testdra-all-usesallresources-hvs5d.driver",
"pool": "worker-5",
"device": "worker-5-device-094"
}
]
},
"nodeSelector": {
"nodeSelectorTerms": [
{
"matchFields": [
{
"key": "metadata.name",
"operator": "In",
"values": [
"worker-5"
]
}
]
}
]
},
"allocationTimestamp": "2026-01-15T14:11:04Z"
}
>
dra_manager.go:280: I0115 15:11:04.408085 18778] scheduler: Device is in flight for allocation device="testdra-all-usesallresources-hvs5d.driver/worker-5/worker-5-device-095" claim="testdra-all-usesallresources-hvs5d/claim-0086"
dra_manager.go:280: I0115 15:11:04.408137 18778] scheduler: Device is in flight for allocation device="testdra-all-usesallresources-hvs5d.driver/worker-5/worker-5-device-096" claim="testdra-all-usesallresources-hvs5d/claim-0165"
default_binder.go:69: I0115 15:11:04.408175 89109] scheduler: Attempting to bind pod to node pod="testdra-all-usesallresources-hvs5d/my-pod-0553" node="worker-5"
dra_manager.go:265: I0115 15:11:04.408264 18778] scheduler: Finished GatherAllocatedState allocatedDevices=<map[string]interface {} | len:2>: {
Initial state: "worker-5-device-094" is in-flight, not in cache
- goroutine #1: starts GatherAllocatedState, copies cache
- goroutine #2: adds to assume cache, removes from in-flight
- goroutine #1: checks in-flight
=> device never seen as allocated
This is the second reason for double allocation of the same device in two
different claims. The other was timing in the assume cache. Both were
tracked down with an integration test (separate commit). It did not fail
all the time, but enough that regressions should show up as flakes.
The distroless-iptables image no longer includes the conntrack binary
as of v0.8.7 (removed in kubernetes/release#4223 since kube-proxy no
longer needs it after kubernetes#126847).
Update the KubeProxy CLOSE_WAIT timeout test to read /proc/net/nf_conntrack
directly instead of using the conntrack command. The file contains the
same connection tracking data and is accessible from the privileged
host-network pod.
Signed-off-by: Davanum Srinivas <davanum@gmail.com>
* DRA: support non-pod references in ReservedFor
Signed-off-by: MohammedSaalif <salifud2004@gmail.com>
* Expand reservation validation comment in syncClaim as suggested by mortent
* Address feedback: rename valid to remaining and remove obsolete TODO
---------
Signed-off-by: MohammedSaalif <salifud2004@gmail.com>
Two test contexts were failing because their AfterEach blocks run even
when BeforeEach skips the test (standard Ginkgo behavior). This caused
resetCRIProxyInjector to fail with "CRI Proxy is undefined".
Switched to using DeferCleanup inside BeforeEach instead, which only
runs if BeforeEach succeeds. This is the same pattern the other test
contexts in this file already use.
Signed-off-by: Davanum Srinivas <davanum@gmail.com>
ImagePullTest was using wait.PollUntilContextCancel which has no
timeout, causing tests to hang for hours when the expected container
state is never reached (e.g., ErrImageNeverPull).
Changed to wait.PollUntilContextTimeout with ContainerStatusRetryTimeout
(5 minutes), matching the pattern used by other tests in the same file.
Signed-off-by: Davanum Srinivas <davanum@gmail.com>
The volume/nfs:1.5 image was updated from CentOS 8 to CentOS Stream 9
in commit 9b47a096f4, but this broke NFS server startup because
NFSv2 is no longer supported in RHEL 9 / CentOS Stream 9.
The `-N 2` flag (disable NFSv2) passed to rpc.mountd and rpc.nfsd
causes them to fail with "2: Unsupported version" since there's no
NFSv2 to disable. This prevented rpc.nfsd from starting, causing
NFS mount timeouts in e2e tests.
Fix by removing the `-N 2` flags. NFSv2 is already unavailable in
CentOS Stream 9, so explicitly disabling it is unnecessary.
Red Hat documentation states: "NFS version 2 (NFSv2) is no longer
supported by Red Hat."
Reference: https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/9/html/managing_file_systems/mounting-nfs-shares_managing-file-systems
Signed-off-by: Davanum Srinivas <davanum@gmail.com>
