This is actually really exciting for me, as it marks the first time during an OpenStack test that I actually was able to fix the problem rather than just deleting the environment.

I did a single-node install of Kolla for messing around. I ran the init-once script and was able to provision an instance without issue. Fan-fucking-tastic.

I went ahead and added 3 more nodes of the same spec just to see if I could. The deployscript errored out at the end for a mariadb issue. I didn’t think a lot of it and just ran the script again. Same failure.

This time, I took a second to look at the red text. It told me to run

kolla-ansible mariadb_recovery

to fix the problem. Fucking sweet, that is really helpful! I ran the script and it said something I didn’t immediately cotton onto “no inventory file passed”, and the script just ended up finishing and not doing anything.

Turns out, the -i flag that I blindly pass my ‘multinode’ file to the other kolla-ansible scripts/playbooks/whatever is for the inventory file, which is my ‘multinode’ file.

kolla-ansible -i multinode mariadb_recovery

ran and apparently did whatever it was supposed to do, because it finished and I re-ran the deploy script and all is well. AWESOME!

This makes me super happy and really ready to take on a more challenging deployment (hyperconverged with ceph!)