rejoin-stack doesn't support RAMCloud

[Shlomo]

steps:
1. Install ramcloud
2. reboot server
3. run rejoin-stack

result:
1. ramcloud services are down (coordinator, server)
2. DF service is down

[Li Ma (nick-ma-z)]

I'll try to figure out how to fix it. BTW, you can run stack.sh instead.
Also, you can specify OFFLINE=True in your local.conf to speed up the process.

[Shlomo (shlominar)]

As far as I know, it's not possible to run stack again, you must run either rejoin or unstack and stack again.

[Li Ma (nick-ma-z)]

In the previous thread, I mean you should run stack.sh after you reboot your server. But I'm not sure rejoin is working as this is what the bug says.

[Shlomo (shlominar)]

Running stack again will erase the whole Openstack environment (networks, VMs, etc.), therefore we have to add support in the rejoin for RAMCloud.

[Li Ma (nick-ma-z)]

The DF service which is down after reboot is related to this bug:
https://bugs.launchpad.net/dragonflow/+bug/1522826

[Shlomo (shlominar)]

Ye, this bug is related to the mentioned bug.

[Li Ma (nick-ma-z)]

The rejoin script is removed in a recent commit in devstack project.
https://github.com/openstack-dev/devstack/commit/082399ef18106bc3820848b6fc352604f0fcf97a