Mark broken / fixed issues
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
MAAS |
Triaged
|
Medium
|
Unassigned | ||
3.5 |
Won't Fix
|
Medium
|
Unassigned | ||
3.6 |
Triaged
|
Medium
|
Unassigned |
Bug Description
maas <user> version read
{
"capabilities": [
"networks-
"static-
"ipv6-
"devices-
"storage-
"network-
"bridging-
"bridging-
"authentica
],
"version": "2.7.0",
"subversion": "8232-g.
}
### Issue 1 ###
Info:
Marking a machine broken and then fixed changes values that should not be changed.
Steps to reproduce:
- Deploy a machine
- Grab machine details: maas <user> machine read <machine> | jq . > orig.json
- Mark as broken
- Mark as fixed
- Grab machine details: maas <user> machine read <machine> | jq . > after-changes.json
- Observe changes: vimdiff orig.json after-changes.json
Example of changed attributes:
"hwe_kernel": "ga-18.04" -> "hwe_kernel": null
"netboot": false -> "netboot": true
"osystem": "ubuntu" -> "osystem": ""
"distro_series": "bionic" -> "distro_series": ""
"current_
"owner": "<user>" -> "owner": null
Expected Results:
None of those attributes should change unless they were adjusted while the machine was marked as broken. This further complicates web UI feedback; owner, OS, and distro are not properly shown.
I believe this bug is related to: https:/
### Issue 2 ###
Info:
I am attempting to change the network interface on a machine to another vlan/subnet.
Steps to reproduce:
REQ: At least two vlan/subnets available...
- Deploy a machine with a single network interface on 1st vlan/subnet (10.11.28.0/24)
- Mark as broken
- Change vlan/subnet to second vlan/subnet (10.11.29.0/24)
- Mark as fixed
- Power on machine
- Observe that the MAAS controller gives the old IP from the 1st vlan/subnet, while the new vlan is set.
Expected Results:
Network interface acquires new IP from the new vlan/subnet
I think this issue relates to MAAS not dropping the machine's IP from the MAAS controller dhcp.leases file. There are probably other locations that MAAS is tracking an IP, but my knowledge of the MAAS database back end / mechanics is nil. It seems as though the controller isn't releasing the IP back to the pool.
My environment looks like:
MAAS Controller:
- IP - 192.168.1.2
- GW - 192.168.1.1
1st vlan/subnet:
10.11.28.0/24
VLAN ID: 200
2nd vlan/subnet:
10.11.29.0/24
VLAN ID: 201
PfSense router with dhcp-relay pointing to 192.168.1.2
MAAS Controller pointing 1st and 2nd vlan/subnet DHCP to MAAS subnet with DHCP Relay
I can provide PCAPs, screenshots, logs if needed. I believe /var/log/syslog or regiond.log is capturing DHCP leases also.
Changed in maas: | |
milestone: | 3.4.0 → 3.4.x |
Changed in maas: | |
milestone: | 3.4.x → 3.5.x |
The bug here seems to be that when you mark a deployed machine as broken, it should still be considered deployed (and broken).