Provide capacity of block-mode RBD volumes #1831
Conversation
nixpanic
added
enhancement
nixpanic
force-pushed
the
rbd/NodeGetVolumeStats
branch
9 times, most recently
from
8741757 to
a4d1031
Compare
nixpanic
force-pushed
the
rbd/NodeGetVolumeStats
branch
5 times, most recently
from
1b3a58a to
73c5826
Compare
|
/retest ci/centos/mini-e2e-helm/k8s-1.19 |
nixpanic
force-pushed
the
rbd/NodeGetVolumeStats
branch
2 times, most recently
from
861dc6e to
089c778
Compare
nixpanic
force-pushed
the
rbd/NodeGetVolumeStats
branch
from
089c778 to
22b5146
Compare
|
/retest all |
nixpanic
force-pushed
the
rbd/NodeGetVolumeStats
branch
3 times, most recently
from
d5a3ec9 to
73d9428
Compare
|
ci/centos/upgrade-tests-cephfs failed again with #1969 ci/centos/mini-e2e/k8s-1.20 failed to bring up the minikube VM |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
Yet again #1969. But the job has been updated, so this is hopefully the last time 🤞 |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
Hit #1969 for cephfs testing. |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
Again failed due to #1969 |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
So annoying, #1969 yet again... We need more debugging of the minikube VM (get libvirt status?) to find out the issue... |
nixpanic
force-pushed
the
rbd/NodeGetVolumeStats
branch
from
bb98a48 to
7ccf9dc
Compare
|
/retest ci/centos/mini-e2e-helm/k8s-1.20 |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
|
Both ci/centos/mini-e2e/k8s-1.20 and ci/centos/mini-e2e-helm/k8s-1.20 failed due to #1969 |
|
/retest ci/centos/mini-e2e-helm/k8s-1.20 |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
|
/retest ci/centos/mini-e2e-helm/k8s-1.20 |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
1 similar comment
|
/retest ci/centos/mini-e2e/k8s-1.20 |
|
/retest ci/centos/mini-e2e-helm/k8s-1.20 |
|
/retest ci/centos/mini-e2e/k8s-1.20 |
|
@Mergifyio rebase |
The CephFS NodeServer should handle the CephFS specific requests. This is not something that the NodeServer for RBD should handle. Signed-off-by: Niels de Vos <ndevos@redhat.com>
The NodeGetVolumeStats procedure can now be used to fetch the capacity of the RBD block-device. By default this is a thin-provisioned device, which means that the capacity is not reserved in the Ceph cluster. This makes it possible to over-provision the cluster. In order to detect the amount of storage used by the RBD block-device (when thin-provisioned), it is required to connect to the Ceph cluster. Unfortunately, the NodeGetVolumeStats CSI procedure does not provide enough parameters to connect to the Ceph cluster and fetch more details about the RBD image. Signed-off-by: Niels de Vos <ndevos@redhat.com>
Test if metrics are available at all. The actual values are a little difficult to validate. BlockMode volumes support metrics since Kubernetes 1.22. See-also: kubernetes/kubernetes#97972 Signed-off-by: Niels de Vos <ndevos@redhat.com>
nixpanic
force-pushed
the
rbd/NodeGetVolumeStats
branch
from
7ccf9dc to
d3f4450
Compare
|
Command
|
Describe what this PR does
Metrics from block-mode volumes are not available from Kubelet. CephFS and RBD-with-filesystem do return metrics. This adds support for the
capacityattribute for block-mode volumes.Related issues
It seems Kubernetes does not call
NodeGetVolumeStatsgot block-mode volumes yet. This will get added through kubernetes/kubernetes#97972 (merged)Without the change in Kubernetes, this can not be tested (and we do not have tests for metrics either yet).
Show available bot commands
These commands are normally not required, but in case of issues, leave any of
the following bot commands in an otherwise empty comment in this PR:
/retest ci/centos/<job-name>: retest the<job-name>after unrelatedfailure (please report the failure too!)
/retest all: run this in case the CentOS CI failed to start/report any testprogress or results