[2.5] Attempting to configure RBAC while MAAS was unable to restart maas-proxy caused internal server error
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
MAAS |
Expired
|
Critical
|
Unassigned |
Bug Description
maas-proxy was failing to start because of this:
Nov 28 12:44:26 rivals.local squid[23937]: Ipc::Mem:
Nov 28 12:44:26 rivals.local sh[23929]: FATAL: Ipc::Mem:
Nov 28 12:44:26 rivals.local systemd[1]: maas-proxy.service: Control process exited, code=exited status=1
Nov 28 12:44:26 rivals.local systemd[1]: maas-proxy.service: Failed with result 'exit-code'.
Which caused the following:
2018-11-28 12:42:29 maasserver.
Traceback (most recent call last):
File "/usr/lib/
File "/usr/lib/
File "/usr/lib/
File "/usr/lib/
--- <exception caught here> ---
File "/usr/lib/
result = result.
File "/usr/lib/
return g.throw(self.type, self.value, self.tb)
File "/usr/lib/
state = yield self.ensureServ
File "/usr/lib/
result = result.
File "/usr/lib/
return g.throw(self.type, self.value, self.tb)
File "/usr/lib/
yield self._performSe
File "/usr/lib/
result = g.send(result)
File "/usr/lib/
raise ServiceActionEr
See "systemctl status maas-proxy.service" and "journalctl -xe" for details.
When trying to make operations agains the API to register RBAC while this was happening, I kept getting the following (with nothing on regiond.log other than the traceback above):
2 roaksoax@
usage: maas [-h] COMMAND ...
optional arguments:
-h, --help show this help message and exit
drill down:
COMMAND
login Log in to a remote API, and remember its description and
logout Log out of a remote API, purging any stored credentials.
list List remote APIs that have been logged-in to.
refresh Refresh the API descriptions of all profiles.
init Initialize controller.
apikey Used to manage a user's API keys. Shows existing keys unless
configauth Configure external authentication.
createadmin Create a MAAS administrator account.
changepassword
admin Interact with http://
test Interact with http://
Internal Server Error
Once I manually fixed the start-up of maas-proxy, attempting to run the command above worked successfully.
Changed in maas: | |
milestone: | none → 2.5.0rc2 |
importance: | Undecided → Critical |
status: | New → Triaged |
Changed in maas: | |
milestone: | 2.5.0rc2 → 2.5.1 |
Changed in maas: | |
milestone: | 2.5.1 → 2.5.2 |
Changed in maas: | |
milestone: | 2.5.2 → 2.5.3 |
Changed in maas: | |
milestone: | 2.5.3 → 2.5.4 |
Changed in maas: | |
milestone: | 2.5.4 → none |
So there are several bugs here.
1) maas-proxy fails to start
2) RBAC was using the same port, 8000, as the proxy
3) The configauth command shows help when something
goes wrong.
I think the real bug that we need to focus on is 1)
2) has already been fixed, RBAC now uses port 5000 by default
3) is not RBAC specific and nothing new. You can reproduce
the same behavior by running the createadmin command twice,
using the same parameters both times.