Closed msdisme closed 1 year ago
Hello Mass Open Cloud Users,
The MOC will be upgrading to OpenShift Container Platform 4.9 on baremetal. Our older OpenShift 3.11 cluster will be decommissioned before the end of this year. If you are interested in continuing to use OpenShift on MOC and would like your projects to be migrated then please fill out this survey.
For further clarifications and inquiries please open a ticket at https://osticket.massopen.cloud
Thanks, MOC Staff
Some minor mods:
The MOC will be upgrading the Intel k-openshift cluster from version OpenShift 3.11 to OpenShift Container Platform version 4.9. OpenShift 4.9 will be running on bare metal cluster benefits/differences from an end user perspective?
Our OpenShift 3.11 cluster will be decommissioned before the end of this year. If you are interested in continuing to use OpenShift on MOC and would like your projects to be migrated then please fill out this survey.
While we expect the process for migrating your containers to be relatively straightforward more details on the differences between the OpenShift versions may be found at this Differences between OpenShift Container Platforms 3 and 4.
_Do we expect people to use, or need to use, the https://docs.openshift.com/container-platform/4.9/migrating_from_ocp_3_to_4/about-mtc-3-4.html? Any other details on what their experience should be like - my sense from discussions with @rob-baron is that most users will just need to redeploy?_
OpenShift 4.9 will be running on bare metal cluster benefits/differences from an end user perspective?
I think the benefits and differences are primarily in the move from 3.11 to 4.9, rather than in the move from OpenStack to bare metal.
While we expect the process for migrating your containers to be relatively straightforward more details on the differences between the OpenShift versions may be found at this Differences between OpenShift Container Platforms 3 and 4.
I don't think that's a useful end user reference. Most (all?) of the differences in that document are primarily of interest to people deploying openshift, e.g.:
It doesn't really address any differences at the container level and won't help people transitioning their workloads from 3.11 to 4.9. I'm not sure I would refer people to that document as part of our announcement; I worry it would just sew confusion.
my sense from discussions with @rob-baron is that most users will just need to redeploy?
I think that will be the easiest method by far, and for people who do run into some sort of problem, we can probably help them on a case-by-case basis.
Another benefit from running on bare metal as opposed to our OpenStack deployment is that it will be higher performance due to the etcd not becoming overwhelmed and constantly having a leader reelection.
Few questions:
It doesn't really address any differences at the container level and won't help people transitioning their workloads from 3.11 to 4.9. I'm not sure I would refer people to that document as part of our announcement; I worry it would just sew confusion.
I agree, we should keep this announcement simple.
we can probably help them on a case-by-case basis.
That's what I think too and they can open a ticket for that.
In my original note, I would get rid of the part that says "projects to be migrated" because I think it sounds like we'll somehow migrate their workloads in the background.
Hello Mass Open Cloud Users,
The MOC will be upgrading to OpenShift Container Platform 4.9 on baremetal. Our older OpenShift 3.11 cluster will be decommissioned before the end of this year. If you are interested in continuing to use OpenShift on MOC please fill out this form here. Once your project is created on the new OpenShift cluster you can start deploying your workloads.
For further clarifications and inquiries please open a ticket at https://osticket.massopen.cloud
Thanks, MOC Staff
I think we are going to wait to send out this email until we have:
ocp-prod
cluster.Once we have those, we can send out an email pointing user to the onboarding system's signup page.
Edit: However, @msdisme still wants to send out an email just so we know how many active users we have so we can plan capacity.
Just some of my thoughts:
Resources currently allocated by the K-Openshift openstack project (that hosts the ocp3 cluster) are:
vCPU: 298 allocated out of 434. RAM: 1 TB out of 11TB ~10TB of Storage.
If a node has 40vCPUs and 128Gigs of memory. then we would need 8 such nodes to be added.
We have available nodes already and we're just waiting on networking equipment; is that correct?
Keep in mind that we have several older projects on k-openshift that we should have cleaned up and that we needed to scale up the cluster towards the end of the semester. After the semester was over, we scaled things back a bit as fewer people were using it.
this is replaced by #816
See comments for more details.