Default storage type from juju add-storage is not in align with the existing one.
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Canonical Juju |
Fix Released
|
High
|
Andrew Wilkins |
Bug Description
Juju version: 2.3-alpha1-
If a charm is deployed with a type-specified storage, e.g.: loop, the type of storage added (the second one) to the unit of that charm, by juju storage, if no type is specified, should be kept same as the first one, in this case it should be loop too.
On AWS bootstrapped controller, if no type is specified for the added storage (the second one), it shows up as ebs. Using charm ceph-osd as an example, the procedures are:
https:/
1. Set credentials, bootstrap a controller on AWS.
2. juju deploy ceph-osd --series xenial --storage osd-devices=loop,1G
3. Wait until the application deployment is finished, up and ready.
4. juju storage
[Storage]
Unit Id Type Pool Provider id Size Status Message
ceph-osd/0 osd-devices/1 block loop volume-1-1 1.0GiB attached
5. juju add-storage ceph-osd/0 osd-devices=1
6. Wait until juju add-storage is complete.
7. juju storage
[Storage]
Unit Id Type Pool Provider id Size Status Message
ceph-osd/0 osd-devices/1 block loop volume-1-1 1.0GiB attached
ceph-osd/0 osd-devices/2 block ebs vol-0345018354e
From output of step 7, the type of added storage should be loop instead of ebs.
Changed in juju: | |
status: | New → Triaged |
importance: | Undecided → High |
Changed in juju: | |
status: | Triaged → In Progress |
assignee: | nobody → Andrew Wilkins (axwalk) |
Changed in juju: | |
status: | In Progress → Fix Committed |
Changed in juju: | |
status: | Fix Committed → Fix Released |
https:/ /github. com/juju/ juju/pull/ 7618