Running make clean makes your host unreachable

[e-minguez]

$ make clean
...
+ virsh net-uuid baremetal
+ virsh net-destroy baremetal
Network baremetal destroyed

+ virsh net-undefine baremetal
Network baremetal has been undefined

+ sudo ifdown baremetal
WARN      : [ifdown] You are using 'ifdown' script provided by 'network-scripts', which are now deprecated.
WARN      : [ifdown] 'network-scripts' will be removed in one of the next major releases of RHEL.
WARN      : [ifdown] It is advised to switch to 'NetworkManager' instead - it provides 'ifup/ifdown' scripts as well.

As it just removes the interfaces, the network connection is lost and I cannot connect back to the host.

[e-minguez]

It happened to me again. This time I was using the ISO:

++ [[ rhel != \r\h\e\l ]]
++ [[ 8 -ne 8 ]]
++ '[' 4085 = 0 ']'
+ virsh net-uuid baremetal
+ virsh net-destroy baremetal
setlocale: No such file or directory
Network baremetal destroyed

+ virsh net-undefine baremetal
setlocale: No such file or directory
Network baremetal has been undefined

+ sudo ifdown baremetal
WARN      : [ifdown] You are using 'ifdown' script provided by 'network-scripts', which are now deprecated.
WARN      : [ifdown] 'network-scripts' will be removed in one of the next major releases of RHEL.
WARN      : [ifdown] It is advised to switch to 'NetworkManager' instead - it provides 'ifup/ifdown' scripts as well.

[e-minguez]

I think this is because in the dev-scripts it removes the interfaces depending on the MANAGE_BR_BRIDGE and MANAGE_PRO_BRIDGE (https://github.com/openshift-metal3/dev-scripts/blob/master/host_cleanup.sh#L26-L37) but here it is done no matter what https://github.com/openshift-kni/install-scripts/blob/master/OpenShift/host_cleanup.sh#L12-L14 https://github.com/openshift-kni/install-scripts/blob/master/OpenShift/host_cleanup.sh#L21-L23 but it seems it shouldn't do it based on the MANAGE_BR_BRIDGE=N parameter that is hardcoded here https://github.com/schmaustech/preflight/blob/master/config_user.j2#L13

[sreichar]

Looking, at this point I don't see much use of make clean.

I think the only thing that currently make sense is clean the ocp sub-directory?

[hardys]

Looking, at this point I don't see much use of make clean.

I think the only thing that currently make sense is clean the ocp sub-directory?

I don't see anything that runs openshift-install destroy cluster, so what cleans up the bootstrap VM in the case of a failed deployment?

Note however that we don't have any support in destroy cluster yet for dealing with the baremetal machines, ref openshift/installer#2005 - I'm not yet sure if that will get done in the 4.3 timeframe, so having some docs or scripted workarounds may be helpful (e.g to at least shut down all existing machines before trying to redeploy)

[sreichar]

@e-minguez have you given this a try?

I would think depending on this result it would change you PR #65 also could address what you mention in issue #71

[e-minguez]

Let me try to create a PR with some more cleaning stuff.