hook failed: "start" with failed to validate CIDR blocks: invalid CIDR address: None/32 error
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
vault-charm |
New
|
Undecided
|
Unassigned |
Bug Description
My Leader of Vault, with an HA config, fails always saying this:
```
unit-vault-29: 08:40:28 ERROR juju.worker.
unit-vault-29: 08:40:28 INFO juju.worker.uniter awaiting error resolution for "start" hook
unit-vault-29: 08:41:47 INFO juju.worker.uniter awaiting error resolution for "start" hook
unit-vault-29: 08:41:48 INFO unit.vault/
unit-vault-29: 08:41:48 ERROR unit.vault/
unit-vault-29: 08:41:48 INFO unit.vault/
unit-vault-29: 08:41:48 INFO unit.vault/
unit-vault-29: 08:41:49 INFO unit.vault/
unit-vault-29: 08:41:49 INFO unit.vault/
unit-vault-29: 08:41:49 INFO unit.vault/
unit-vault-29: 08:41:50 INFO unit.vault/
unit-vault-29: 08:41:50 INFO unit.vault/
unit-vault-29: 08:41:50 INFO unit.vault/
unit-vault-29: 08:41:50 ERROR unit.vault/
Traceback (most recent call last):
File "/var/lib/
bus.
File "/var/lib/
_invoke(
File "/var/lib/
handler.
File "/var/lib/
self.
File "/var/lib/
approle_id = vault.configure
File "/var/lib/
client.
File "/var/lib/
return self._adapter.
File "/var/lib/
return self.request(
File "/var/lib/
utils.
File "/var/lib/
raise exceptions.
hvac.exceptions
* failed to validate CIDR blocks: invalid CIDR address: None/32
```
The vault juju config are this:
```
vault:
charm: cs:vault
num_units: 3
to:
- lxd:0
- lxd:1
- lxd:2
options:
vip: xxxxxxxxx xxxxxxxxxx
bindings:
"": *internal
ha: *admin
vault-
charm: cs:mysql-router
bindings:
db-router: *internal
shared-db: *internal
relations:
- - vault-mysql-
- mysql-innodb-
- - vault:shared-db
- vault-mysql-
- - vault:etcd
- etcd:db
- - ha-vault:ha
- vault:ha
- - nova-compute-
- vault:secrets
- - vault:certificates
- neutron-
- - ovn-central:
- vault:certificates
- - ovn-chassis:
- vault:certificates
```
My rev of Vault is 54. Any hint welcomed!
Changed in vault-charm: | |
status: | New → Invalid |
status: | Invalid → New |
I've been seeing this with vault revision 82 (channel 1.7/stable).
When running the hooks, I added some debug code and see that two relations in the requests list have ingress_ address= None; Units ceph-osd/2 and ceph-osd/8. All the other ceph-osd units seem to have worked as expected.
I was able to work around the issue by using request[ 'access_ address' ] if request[ 'ingress_ address' ] is None. (Simple patch attached.)
I am not sure if this is a good general solution but in my situation, where the ingress_address is set, it's the same as the access_address.