Updated commands to `oc adm drain node` #5709

ahardin-rh · 2017-10-13T19:09:13Z

Addresses #5613

@pravisankar Can you please check the syntax here?

ahardin-rh · 2017-11-08T16:52:40Z

pravisankar · 2017-11-09T05:01:22Z

admin_guide/manage_nodes.adoc

@@ -245,22 +245,19 @@ To list pods that will be migrated without actually performing the evacuation,
 use the `--dry-run` option:

 ----
-$ oadm manage-node <node1> <node2> \
-    --evacuate --dry-run [--pod-selector=<pod_selector>]
+$ oadm drain <node1> <node2>  --dry-run [--pod-selector=<pod_selector>]


--pod-selector is not a valid option for oadm drain.
Supported options:

Usage: oadm drain NODE [options] Examples: # Drain node "foo", even if there are pods not managed by a ReplicationController, ReplicaSet, Job, DaemonSet or StatefulSet on it. $ oadm drain foo --force # As above, but abort if there are pods not managed by a ReplicationController, ReplicaSet, Job, DaemonSet or StatefulSet, and use a grace period of 15 minutes. $ oadm drain foo --grace-period=900 Options: --delete-local-data=false: Continue even if there are pods using emptyDir (local data that will be deleted when the node is drained). --dry-run=false: If true, only print the object that would be sent, without sending it. --force=false: Continue even if there are pods not managed by a ReplicationController, ReplicaSet, Job, DaemonSet or StatefulSet. --grace-period=-1: Period of time in seconds given to each pod to terminate gracefully. If negative, the default value specified in the pod will be used. --ignore-daemonsets=false: Ignore DaemonSet-managed pods. --timeout=0s: The length of time to wait before giving up, zero means infinite

ahardin-rh · 2017-12-06T19:34:14Z

@pravisankar Thanks for your guidance. This is now updated. PTAL.

pravisankar

I'm not sure if this doc holds true for enterprise 3.5, 3.6, 3.7 and 3.9.
At least latest origin code (3.9) currently does not show --dry-run option but 3.9 is not released yet. Currently k8s rebase is in progress (openshift/origin#17576) and we may have --pod-selector and selector options (openshift/origin#17616) and also --dry-run option (openshift/origin#16333).
I would recommend holding off publishing this change for 3.9 until we have final build ready and make sure all the oadm drain flags are same in 3.5, 3.6 and 3.7.

pravisankar · 2017-12-07T22:33:01Z

admin_guide/manage_nodes.adoc

 ----

+If `dry-run` is set to `true`, you only print the object that would be sent,
+without sending it.
+
 To actually evacuate all or selected pods on one or more nodes:


To actually evacuate all pods on one or more nodes: (currently there is no option to select pods)

ahardin-rh · 2017-12-11T16:45:44Z

@pravisankar Okay, thanks. I will check back in later on this.

miminar · 2017-12-14T12:24:45Z

admin_guide/manage_nodes.adoc

@@ -245,27 +245,45 @@ To list pods that will be migrated without actually performing the evacuation,
 use the `--dry-run` option:

 ----
-$ oadm manage-node <node1> <node2> \
-    --evacuate --dry-run [--pod-selector=<pod_selector>]
+$ oadm drain <node1>  --dry-run=true


oadm has been removed: openshift/origin#17396
it's being re-added already but with a fat decprecated warning though
oc adm is preferred

miminar · 2017-12-14T12:25:03Z

👍

I've just received complaints about this from a customer. I'm pleased to see it being fixed.

ghost · 2018-02-13T18:56:23Z

@ahardin-rh Any update on this? I have a bug that is related: https://bugzilla.redhat.com/show_bug.cgi?id=1477380

Cc: @vikram-redhat

ahardin-rh · 2018-02-19T20:28:11Z

@tmorriso-rh Per discussion earlier in this PR, we decided to hold off on merging this until there is a final build ready so that we can confirm that oc adm drain flags are the same in 3.5, 3.6, and 3.7.
@pravisankar Any updates on this?

pravisankar · 2018-02-20T16:56:39Z

drain is an upstream k8s CLI command, corresponding kube docs already available but depending on when we do k8s rebase in openshift, available CLI options may vary. I'm guessing we are also documenting these commands to exactly show available command options for a particular release.

This is how drain command was evolved:
Options available in 3.5 and 3.6 release:

      --delete-local-data=false: Continue even if there are pods using emptyDir (local data that will be deleted when
the node is drained).
      --force=false: Continue even if there are pods not managed by a ReplicationController, ReplicaSet, Job, DaemonSet
or StatefulSet.
      --grace-period=-1: Period of time in seconds given to each pod to terminate gracefully. If negative, the default
value specified in the pod will be used.
      --ignore-daemonsets=false: Ignore DaemonSet-managed pods.
      --timeout=0s: The length of time to wait before giving up, zero means infinite

3.7 release:

In addition to options in 3.6
--dry-run=false: If true, only print the object that would be sent, without sending it.

3.8 release:

In addition to options in 3.7
--pod-selector='': Label selector to filter pods on the node
option --dry-run was removed, not sure why that was done

3.9 release:

In addition to options in 3.8
[option '--dry-run' was re-introduced]
--dry-run=false: If true, only print the object that would be sent, without sending it.
 -l, --selector='': Selector (label query) to filter on

pravisankar

We have enough information now to doc this command. Provided needed details

ahardin-rh · 2018-02-20T21:28:22Z

@pravisankar Thanks for your guidance! PTAL. This PR addresses 3.5 and 3.6. I will open companion PRs for 3.7 and 3.9 to add in the additional details for each. We are not releasing docs for 3.8, but it's interesting seeing the evolution. Thanks again for all of your help in getting this moved along!

pravisankar

Minor nit otherwise LGTM

pravisankar · 2018-02-20T22:21:00Z

admin_guide/manage_nodes.adoc

@@ -241,31 +241,49 @@ controller] can be evacuated; the replication controllers create new pods on
 other nodes and remove the existing pods from the specified node(s). Bare pods,
 meaning those not backed by a replication controller, are unaffected by default.

-To list pods that will be migrated without actually performing the evacuation,
-use the `--dry-run` option:
+To actually evacuate all or selected pods on one or more nodes:


To actually evacuate one or more nodes:
[Selecting pods needs --pod-selector and this option is not available in 3.5/3.6]

@pravisankar Updated. Thank you!

ahardin-rh · 2018-02-21T02:19:34Z

Updated per feedback. I will merge this and build off of it for 3.7 and 3.9 docs in separate PRs. Revision histories for each will be applied in the cherry-pick PRs.

ahardin-rh · 2018-02-21T02:24:37Z

/cherrypick enterprise-3.5

openshift-cherrypick-robot · 2018-02-21T02:24:44Z

@ahardin-rh: new pull request created: #7874

In response to this:

/cherrypick enterprise-3.5

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

ahardin-rh · 2018-02-21T02:25:03Z

/cherrypick enterprise-3.6

openshift-cherrypick-robot · 2018-02-21T02:25:12Z

@ahardin-rh: new pull request created: #7875

In response to this:

/cherrypick enterprise-3.6

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

ahardin-rh added branch/enterprise-3.4 branch/enterprise-3.5 branch/enterprise-3.6 labels Oct 13, 2017

ahardin-rh added this to the Next Release milestone Oct 13, 2017

ahardin-rh self-assigned this Oct 13, 2017

pravisankar reviewed Nov 9, 2017

View reviewed changes

ncbaratta added the peer-review-done Signifies that the peer review team has reviewed this PR label Nov 9, 2017

ahardin-rh force-pushed the oadm-drain branch from c66c5a2 to 2b0f879 Compare December 6, 2017 19:33

openshift-ci-robot added the size/M Denotes a PR that changes 30-99 lines, ignoring generated files. label Dec 6, 2017

ahardin-rh added branch/enterprise-3.7 branch/enterprise-3.9 and removed branch/enterprise-3.4 labels Dec 6, 2017

pravisankar reviewed Dec 7, 2017

View reviewed changes

miminar reviewed Dec 14, 2017

View reviewed changes

vikram-redhat modified the milestones: Next Release, Staging Jan 8, 2018

openshift-bot added the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label Feb 17, 2018

pravisankar suggested changes Feb 20, 2018

View reviewed changes

ahardin-rh force-pushed the oadm-drain branch from 2b0f879 to eca6ab2 Compare February 20, 2018 21:23

openshift-bot removed the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label Feb 20, 2018

ahardin-rh removed the branch/enterprise-3.7 label Feb 20, 2018

ahardin-rh removed the branch/enterprise-3.9 label Feb 20, 2018

ahardin-rh changed the title ~~Updated commands to oadm drain node~~ Updated commands to oc adm drain node Feb 20, 2018

pravisankar suggested changes Feb 20, 2018

View reviewed changes

ahardin-rh force-pushed the oadm-drain branch from eca6ab2 to 5d9f491 Compare February 21, 2018 02:01

Updated commands to oadm drain NODE

5d9f491

ahardin-rh added the to_followup label Feb 21, 2018

ahardin-rh merged commit 9dc4af3 into openshift:master Feb 21, 2018

openshift-cherrypick-robot mentioned this pull request Feb 21, 2018

[enterprise-3.5] Updated commands to oc adm drain node #7874

Closed

openshift-cherrypick-robot mentioned this pull request Feb 21, 2018

[enterprise-3.6] Updated commands to oc adm drain node #7875

Closed

This was referenced Feb 21, 2018

Added --dry-run option for oc adm drain for 3.7 #7880

Merged

Added oc adm drain options for 3.9 #7881

Merged

ahardin-rh removed the to_followup label Feb 21, 2018

Updated commands to oc adm drain node #5709

Updated commands to oc adm drain node #5709

Uh oh!

Conversation

ahardin-rh commented Oct 13, 2017

Uh oh!

ahardin-rh commented Nov 8, 2017

Uh oh!

pravisankar Nov 9, 2017

Choose a reason for hiding this comment

Uh oh!

ahardin-rh commented Dec 6, 2017

Uh oh!

pravisankar left a comment

Choose a reason for hiding this comment

Uh oh!

pravisankar Dec 7, 2017

Choose a reason for hiding this comment

Uh oh!

ahardin-rh commented Dec 11, 2017

Uh oh!

miminar Dec 14, 2017

Choose a reason for hiding this comment

Uh oh!

miminar commented Dec 14, 2017

Uh oh!

ghost commented Feb 13, 2018

Uh oh!

ahardin-rh commented Feb 19, 2018

Uh oh!

pravisankar commented Feb 20, 2018

Uh oh!

pravisankar left a comment

Choose a reason for hiding this comment

Uh oh!

ahardin-rh commented Feb 20, 2018

Uh oh!

pravisankar left a comment

Choose a reason for hiding this comment

Uh oh!

pravisankar Feb 20, 2018

Choose a reason for hiding this comment

Uh oh!

ahardin-rh Feb 21, 2018

Choose a reason for hiding this comment

Uh oh!

ahardin-rh commented Feb 21, 2018

Uh oh!

ahardin-rh commented Feb 21, 2018

Uh oh!

openshift-cherrypick-robot commented Feb 21, 2018

Uh oh!

ahardin-rh commented Feb 21, 2018

Uh oh!

openshift-cherrypick-robot commented Feb 21, 2018

Uh oh!

Uh oh!

Updated commands to `oc adm drain node` #5709

Updated commands to `oc adm drain node` #5709