redhat-cop / ocp4-helpernode

This playbook helps set up an "all-in-one" node, that has all the infrastructure/services in order to install OpenShift 4.
342 stars 303 forks source link

ansible-playbook failure: Could not find the requested service helper-tftp: host #269

Closed steve-delillo closed 2 years ago

steve-delillo commented 2 years ago

Hello. There must have been a change to the playbook, recently, because I have ran this playbook multiple times without seeing the following error. Can anyone tell me why we get the following error on a newly created RHEL8.3 server? Also, I cloned the repository, today, Jan 31. I had ran the same playbook with the exact same vars.yaml file that I used a few weeks ago (except I added ipi: false) and did not hit this failure. Usually, the playbook creates and tries to start the helper-tftp service, but no longer. This is for a bare-metal install.

image

~# systemctl status helper-tftp.service Unit helper-tftp.service could not be found.

Here, we can see that the playbook did not put either of the tftp start scripts where they should be: ~# ls /usr/local/bin/start-tftp.sh ls: cannot access '/usr/local/bin/start-tftp.sh': No such file or directory

~# ls /etc/systemd/system/helper-tftp.service ls: cannot access '/etc/systemd/system/helper-tftp.service': No such file or directory

ftp service did start, though: ~# systemctl status tftp.service ● tftp.service - Tftp Server Loaded: loaded (/usr/lib/systemd/system/tftp.service; indirect; vendor preset: disabled) Active: active (running) since Mon 2022-01-31 15:49:27 EST; 29s ago Docs: man:in.tftpd Main PID: 69220 (in.tftpd) Tasks: 1 (limit: 3295900) Memory: 248.0K CGroup: /system.slice/tftp.service └─69220 /usr/sbin/in.tftpd -s /var/lib/tftpboot

~# cat vars.yaml

disk: sda helper: name: "helper" ipaddr: "192.100.229.190" networkifacename: "ens2f1" dns: domain: "ocpexample.com" clusterid: "mycluster" forwarder1: "9.11.227.2" forwarder2: "9.0.128.5" dhcp: router: "192.100.229.1" bcast: "192.100.229.255" netmask: "255.255.255.0" poolstart: "192.100.229.191" poolend: "192.100.229.195" ipid: "192.100.229.0" netmaskid: "255.255.255.0" bootstrap: name: "bootstrap" ipaddr: "192.100.229.191" macaddr: "b8:ce:f6:78:2d:6f" masters:

christianh814 commented 2 years ago

Are you still having this issue?

This was actually a workaround for a bug in RHEL 7 ...so I'm not even sure this is needed anymore. I'll do some testing on my end.

In the meantime, does this happen all the time?

steve-delillo commented 2 years ago

Hi Christian. I somehow got around it. Tried a few things, but I think that this is what fixed it.

In the ocp4-helpernode/vars/main.yml, I changed both of these to false. When I first hit the issue, I only had ipi as false. This is a baremetal install, but the "baremetal" value may mean something else. ipi: false baremetal: false

christianh814 commented 2 years ago

Ah it looks like it does need to be set on both. Wonder if it should be an or

https://github.com/redhat-cop/ocp4-helpernode/blob/main/tasks/main.yml#L226

Not 100% sure because I don't remember why baremetal is even needed if the original purpose of this playbook was specifically for Baremetal.

Glad you got it working!

christianh814 commented 2 years ago

Closing this issue as this was fixed in #264 should merge into main soon.