[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Newbie [Centos 7.7] - OKD 3.11: Host network already using 172.16.x.x



Hi Denny,

Cluster network and portal network should not overlap with physical network.

You can setup networking of OKD in inventory file (osm_cluster_network_cidr and openshift_portal_net).

For example set osm_cluster_network_cidr to 10.128.0.0/16 and openshift_portal_net 172.30.0.0/16 which should avoid 172.16.0.0/16 network

More details can be found in
https://docs.openshift.com/container-platform/3.11/install/configuring_inventory_file.html (Table 2.)


Uz pozdrav,

Vladimir Remenar
Aplikativni razvoj
Application Development
DevOps inženjer / DevOps Engineer



Strongest Banking Brand in the World 2018
, Brand Finance
Best Bank in CEE 2018
, The Banker magazine


Sberbank d.d.

Varšavska 9, 10000 Zagreb, Hrvatska
T +385 1 4801 333
F +385 1 4801 365
vladimir remenar sberbank hr
https://www.sberbank.hr

Commercial Court in Zagreb, MBS 080126398
https://www.sberbank.hr/disclaimer





From:        "Denny Fuchs" <linuxmail 4lin net>
To:        users lists openshift redhat com
Date:        04.02.2020 13:12
Subject:        Newbie [Centos 7.7] - OKD 3.11:  Host network already using 172.16.x.x
Sent by:        users-bounces lists openshift redhat com




Hello,

I'm pretty new to OKD and CentOs and dropped my test Kubernetes cluster.
I followed the small HOWTO from:
https://github.com/arunvdsharma/openshift-centos.git
This is a small wrapper for the Ansible OKD.

My Test setup:

* 1 x Master
* 2 x Infra
* 2 x Compute


Ansible runs more (or less) without big issues over the nodes and brings
the most parts online: except the network.
I have a lot of messages like:

Feb  4 11:34:48 fra-test-node-master origin-node: W0204 11:34:48.663409  
 33214 cni.go:172] Unable to update cni config: No networks found in
/etc/cni/net.d


but I've found it on many sites, and most of them says .. its pretty
normal.

Also:

94a0d2e374db4a71b81896)"), skipping: failed to "StartContainer" for
"controllers" with CrashLoopBackOff: "Back-off 5m0s restarting failed
container=controllers
pod=master-controllers-fra-test-node-master.dev.ina_kube-system(3210f8756194a0d2e374db4a71b81896)

or:

Feb  4 11:36:16 fra-test-node-master origin-node: exec openshift start
master controllers --config=/etc/origin/master/master-config.yaml
--listen=https://0.0.0.0:8444 --loglevel=${DEBUG_LOGLEVEL:-2}

Feb  4 11:36:16 fra-test-node-master origin-node: ] WorkingDir: Ports:[]
EnvFrom:[] Env:[] Resources:{Limits:map[] Requests:map[]}
VolumeMounts:[{Name:master-config ReadOnly:false
MountPath:/etc/origin/master/ SubPath: MountPropagation:<nil>}
{Name:master-cloud-provider ReadOnly:false
MountPath:/etc/origin/cloudprovider/ SubPath: MountPropagation:<nil>}
{Name:signature-import ReadOnly:false
MountPath:/etc/containers/registries.d/ SubPath: MountPropagation:<nil>}
{Name:kubelet-plugins ReadOnly:false
MountPath:/usr/libexec/kubernetes/kubelet-plugins SubPath:
MountPropagation:0xc4221fb160} {Name:master-pki ReadOnly:false
MountPath:/etc/pki SubPath: MountPropagation:<nil>} {Name:host-localtime
ReadOnly:false MountPath:/etc/localtime SubPath:
MountPropagation:<nil>}] VolumeDevices:[]
LivenessProbe:&Probe{Handler:Handler{Exec:nil,HTTPGet:&HTTPGetAction{Path:healthz,Port:8444,Host:,Scheme:HTTPS,HTTPHeaders:[],},TCPSocket:nil,},InitialDelaySeconds:0,TimeoutSeconds:1,PeriodSeconds:10,SuccessThreshold:1,FailureThreshold:3,}
ReadinessProbe:nil Lifecycle:nil
TerminationMessagePath:/dev/termination-log
TerminationMessagePolicy:File ImagePullPolicy:IfNotPresent
SecurityContext:&SecurityContext{Capabilities:nil,Privileged:*true,SELinuxOptions:nil,RunAsUser:nil,RunAsNonRoot:nil,ReadOnlyRootFilesystem:nil,AllowPrivilegeEscalation:nil,RunAsGroup:nil,}
Stdin:false StdinOnce:false TTY:false} is dead, but RestartPolicy says
that we should restart it.

Feb  4 11:36:16 fra-test-node-master origin-node: I0204 11:36:16.863315  
 33214 kuberuntime_manager.go:757] checking backoff for container
"controllers" in pod
"master-controllers-fra-test-node-master.dev.ina_kube-system(3210f8756194a0d2e374db4a71b81896)"

Feb  4 11:36:16 fra-test-node-master origin-node: I0204 11:36:16.863616  
 33214 kuberuntime_manager.go:767] Back-off 5m0s restarting failed
container=controllers
pod=master-controllers-fra-test-node-master.dev.ina_kube-system(3210f8756194a0d2e374db4a71b81896)


But however ... I think the big problem is, that the host network, where
I installed my test cluster, is 172.16.x.x which is the default for OKD.
I can't change the company network and also later the production network
..

So the inventory.ini looks like:

[OSEv3:children]
masters
nodes
etcd

[masters]
172.16.0.86

[etcd]
172.16.0.86
172.16.0.74
172.16.0.75
172.16.0.76
172.16.0.77

[nodes]
172.16.0.86 openshift_node_group_name='node-config-master'
172.16.0.74 openshift_node_group_name='node-config-infra'
172.16.0.75 openshift_node_group_name='node-config-infra'
172.16.0.76 openshift_node_group_name='node-config-compute'
172.16.0.77 openshift_node_group_name='node-config-compute'

...

[OSEv3:vars]
...
openshift_public_hostname=console.dev.ina
openshift_master_default_subdomain=apps.dev.ina
...

So my newbie question is: how to get around that or change the
addresses, which OKD can use, without getting routing issues ...

I think, I misunderstood the openshift_portal_net param, because after
change it ... and redeploy ... I had much more trouble:

openshift_portal_net=192.168.99.0/24


It would be great, of someone could help me :-)

cu denny

_______________________________________________
users mailing list
users lists openshift redhat com
http://lists.openshift.redhat.com/openshiftmm/listinfo/users




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]