Re: specifying storage class for metrics and logging

Hi. On 17 April 2018 at 12:17, Tim Dudgeon <tdudgeon...@gmail.com> wrote: > So if you are using dynamic provisioning the only option for logging is > for the default StorageClass to be set to what is needed? > > On 17/04/18 11:12, Per Carlson wrote: > > Thi

Version 3.8 and 3.9 confusion

Hi. I just noticed a message on the dev mailinglist: > We've branched master for release-3.8 and created a v3.9.0-alpha.0 tag. > This is because 3.8 is a "skip" release where we'll only do an internal > data upgrade and then go from 3.7 to 3.9 directly. Does that mean there will never be a

Fwd: LDAP bindPassword in Ansible inventory

And to the list as well... ​--8<--​ If having the password in clear text in master-coinfig.yaml is OK, you could also use this trick: openshift_master_identity_providers=[{'name':'OpenID', 'kind':'OpenIDIdentityProvider', 'clientSecret':"{{ lookup('file','/path/to/secret') }}"}] This way

Re: Scheduler policy in 3.6 (default and custom at install time)

Hi. On 12 August 2017 at 18:39, Avesh Agarwal <avaga...@redhat.com> wrote: > > > On Sat, Aug 12, 2017 at 11:59 AM, Avesh Agarwal <avaga...@redhat.com> > wrote: > >> >> >> On Fri, Aug 11, 2017 at 2:28 AM, Per Carlson <pe...@hemmop.com> wrote: >

Scheduler policy in 3.6 (default and custom at install time)

Hi. We are in the process of rebuilding a cluster with a new topology, and I'm trying to fit the node labels to a scheduler policy, and would like to base the policy on the default one. I've searched both the openshift/origin and and openshift/openshift-ansible repos on GitHub without finding

Re: Health check of API-server and HA-proxy

​Hi Alexandar. ​ > > There is a /healthz endpoint in HA-proxy on port 1936, but it isn't > > exposed outside the cluster and requires a password (which is unique > > per dc). What else could be used? We would like to stay as close as > > "stock configuration" as possible to reduce technical

Re: Strange TLS error from oc

Hi Jordan. On 21 July 2017 at 18:58, Jordan Liggitt wrote: > Looks like an invalid cert. See https://github.com/golang/go/issues/15407 > for details. > Figured that out by doing a packet capture. Wireshark did complain about the invalid length of the time string.

Strange TLS error from oc

When trying to login to a (new) OCP 3.5 cluster, oc emits this error: error: tls: failed to parse certificate from server: parsing time "20171001Z" as "20060102150405Z0700": cannot parse "Z" as "05" ​The certificate (which I got via "openssl s_client") has got: ​ Validity Not Before: Jul

Re: Method to move a single or multiple pods to a different node?

On 12 July 2017 at 00:50, G. Jones wrote: > That’s just it, the masters were unschedulable. During the outage wer > restarted the masters and nodes but the nodes wouldn’t come online. While > we were working on getting the nodes up the pods had been restarted on the >

Re: Method to move a single or multiple pods to a different node?

Hi. On 8 July 2017 at 21:45, G. Jones wrote: > I’ve got an Origin 1.5 environment where, during an outage with my nodes, > pods got relocated to my masters fairly randomly. I need to clean it up and > get the pods back where I want them but have not yet found a way to

Re: Increase HeapSize on registry.access.redhat.com/jboss-webserver-3/webserver30-tomcat8-openshift:1.2-12

Hi. On 5 December 2016 at 09:39, Den Cowboy wrote: > Thanks for your response Per. > > I can confirm it's dynamically. We increased our resources and the > heapsize increased for our tomcat. > > When we want to increase the heap_size with the size of our environment >

Re: Increase HeapSize on registry.access.redhat.com/jboss-webserver-3/webserver30-tomcat8-openshift:1.2-12

Hi Den. The heap allocation is done dynamically. When the image starts it runs /opt/webserver/bin/launch.sh. In this file you will find MAX_HEAP=`get_heap_size` if [ -n "$MAX_HEAP" ]; then CATALINA_OPTS="$CATALINA_OPTS -Xms${MAX_HEAP}m -Xmx${MAX_HEAP}m" fi The function

Re: Best way to delete images (local and in registry)

What about cleaning the local docker storage? We have had several "incidents" were deployments of a POD failed due to "out of disk" errors on the node. For example on an infrastructure node: $ sudo docker info Containers: 8 Images: 23 Storage Driver: devicemapper Pool Name:

Timezone for PODs

Hi. It looks like PODs are logging with a different timezone than the host system. This is of course cosmetic, but nevertheless annoying. ​The host is configured with CET (UTC+1)​: [root@infra201 ~]# date Mon Feb 22 14:51:04 CET 2016 But the docker-registry is using (UTC-5): [root@infra201

Re: MTU on tun0

Hi David. On 12 February 2016 at 09:45, David Strejc wrote: > This is a ansible installer problem. > > I had similar issue with my instalation and this is really important. > Without MTU set to 1450 there is > a problem with traffic between nodes. I've spent four hours

MTU on tun0

Hi. ​We are seeing some strange packet traces on the nodes, and we suspect that it might be a MTU-issue.​ According to the documentation ( https://docs.openshift.com/enterprise/3.1/install_config/configuring_sdn.html#configuring-the-pod-network-on-nodes) a "mtu" parameter in node-config.yaml