enable hwmon for sensor collection for bare metal clusters #971
Conversation
dhellmann
force-pushed
the
enable-hwmon
branch
from
25034f4
to
4eb7cd5
Compare
| @@ -23,7 +23,6 @@ spec: | |||
| - --path.sysfs=/host/sys | |||
| - --path.rootfs=/host/root | |||
| - --no-collector.wifi | |||
| - --no-collector.hwmon | |||
There was a problem hiding this comment.
This modifies the generate assets directly, you need to modify the jsonnet file to remove this, see this guide how to then generate https://github.com/openshift/cluster-monitoring-operator/blob/master/CONTRIBUTING.md
There was a problem hiding this comment.
Hmm, this was the only file in this repository that had that string in it. Where is this file coming from upstream?
There was a problem hiding this comment.
PR that removed hwmon from enabled collectors in upstream repository: prometheus-operator/kube-prometheus#381
| @@ -7,6 +7,7 @@ | |||
| - Adjusted NodeClockNotSynchronising, NodeNetworkReceiveErrs, and NodeNetworkTransmitErrs alerts. | |||
| - [#962](https://github.com/openshift/cluster-monitoring-operator/pull/962) Enable namespace by pod and pod total networking Grafana dashboards. | |||
| - [#959](https://github.com/openshift/cluster-monitoring-operator/pull/959) Remove memory limits from prometheus-config-reloader in user workload monitoring | |||
| - [#971](https://github.com/openshift/cluster-monitoring-operator/pull/971) Enable `hwmon` in node-exporter for hardware sensor data collection | |||
There was a problem hiding this comment.
I remember us disabling not urgent collectors due to high cardinality, I am wondering how many new series this brings in on each cluster?
There was a problem hiding this comment.
The number of series depends on the number of sensors visible on the node, so it's hard to pin that down. On my dev system, I only get some temperature sensors. On other hosts, we should see more.
There was a problem hiding this comment.
@dhellmann I am wondering is this enablement part of some concrete initiative?
There was a problem hiding this comment.
Here's some example output from one host: https://paste.centos.org/view/e2d24856
That includes temperature data, but that host is apparently not exposing fan speed or other sensors.
dhellmann
force-pushed
the
enable-hwmon
branch
from
4eb7cd5
to
e1c6c72
Compare
|
/retest |
|
I've updated the PR with a jsonnet expression to remove the flag from the upstream version of the deployment settings. Regenerating the assets modified manifests/0000_50_cluster-monitoring-operator_02-role.yaml but it looks like that has to do with something not sorting output consistently. I can manually remove that change, but thought I should include all of the output of |
dhellmann
force-pushed
the
enable-hwmon
branch
from
e1c6c72
to
f83ed78
Compare
|
/retest |
2 similar comments
|
/retest |
|
/retest |
dhellmann
force-pushed
the
enable-hwmon
branch
2 times, most recently
from
2a3a858
to
d745180
Compare
|
/test generate |
|
@dhellmann this lgtm, but could you please rebase? |
openshift-ci-robot
added
the
needs-rebase
Enable the hwmon data collection so that hardware telemetry like CPU temperature and fan speeds are available for bare metal clusters. Signed-off-by: Doug Hellmann <dhellmann@redhat.com>
dhellmann
force-pushed
the
enable-hwmon
branch
from
d745180
to
00db9df
Compare
openshift-ci-robot
removed
the
needs-rebase
|
Rebased to resolve the changelog merge conflict. |
|
/lgtm |
1 similar comment
|
/lgtm |
|
/hold |
openshift-ci-robot
added
the
do-not-merge/hold
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: dhellmann, s-urbaniak The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
openshift-ci-robot
added
the
approved
|
just holding to verify that @lilic's comments have been addressed. |
|
/retest |
1 similar comment
|
/retest |
|
/hold cancel lgtm 👍 |
openshift-ci-robot
removed
the
do-not-merge/hold
|
/retest Please review the full test history for this PR and help us cut down flakes. |
1 similar comment
|
/retest Please review the full test history for this PR and help us cut down flakes. |
Enable the hwmon data collection so that hardware telemetry like CPU
temperature and fan speeds are available for bare metal clusters.