almostobvious commented on issue #11024:
URL: https://github.com/apache/cloudstack/issues/11024#issuecomment-2972022385
Ok so it only took me ~10 hours to add a host, but I got there.
1. I had to use cmk to clean up previous attempts at primary storage etc for
now non-existent host. This was a bit confusing as I thought it was the cause
of issues (but it wasn't). Still, this can't be properly done from the web ui
because it lacks a whole raft of options - e.g. you can't put storagepool to
maintenance.
In my case I had to do:
```
enable storagemaintenance id=<old / orphaned storage id>
delete storagepool id=22594f15-9678-347e-9627-0f935db15e9f
```
Which finally allowed me to get rid of orphaned storage.
2. I uninstalled cloudstack-agent. Then cleaned up all config by finding and
deleting a whole bunch of stuff that gets left behind after uninstall (and I
was suspecting of causing havoc, but it really didn't). I used following (and
some good judgement) to find what to delete
```
find /usr/ -name "*libvirt*"
find /etc/ -name "*libvirt*"
find /var/ -name "*libvirt*"
find /usr/ -name "*cloudstack*"
find /etc/ -name "*cloudstack*"
find /var/ -name "*cloudstack*"
```
3. reboot, and then force installation of previous agent version that worked
for me on RHEL/Rocky 9.5:
`yum install cloudstack-agent-4.20.0.0-1`
4. Now I went back and followed Quick install guide KVM config section here
https://docs.cloudstack.apache.org/en/latest/quickinstallationguide/qig.html#kvm-configuration
5. Now make sure nfs-server is running and set up correctly:
```
// verify /etc/exports as follows:
/exports/secondary *(rw,async,no_root_squash,no_subtree_check)
/exports/primary *(rw,async,no_root_squash,no_subtree_check)
// ensure correct ownership
chown nobody:nobody /exports/primary/
chown nobody:nobody /exports/secondary/
// nfs server check
systemctl enable --now nfs-server
exportfs -r
```
6. Start / enable agent
```
systemctl enable cloudstack-agent
systemctl start cloudstack-agent
```
7. Now I used CMK to create host:
`add host zoneid=<id> clusterid=<id> hypervisor=KVM username=root
password=<pass> podid=<podid> url=http://<new host ip>`
That took a little time, but succeeded.
I am now pretty sure the issue is with the new agent build 4.20.1.0-1
(whether or not it's down to the spurious ubuntu / paleo stuff it seems to be
doing on RHEL I can't tell).
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
