Lookup SCSI Controller Device Type from Hardware [Depends manageiq-gems-pending/148] #51
Conversation
agrare
changed the title
agrare
force-pushed
the
bz_1445874_vm_reconfigure_scsi_controller
branch
from
46921d1 to
e5adccd
Compare
|
@blomquisg can you take a look? |
| @@ -92,16 +92,24 @@ def build_config_spec(options) | |||
|
|
|||
| if options[:disk_remove] || options[:disk_add] | |||
| with_provider_object do |vim_obj| | |||
| options[:disk_remove].each { |d| remove_disk_config_spec(vim_obj, vmcs, d) } if options[:disk_remove] | |||
| add_disks(vim_obj, vmcs, options[:disk_add]) if options[:disk_add] | |||
| hardware = vim_obj.send(:getHardware) | |||
There was a problem hiding this comment.
Do you have to send here? Could this just be vim_obj.getHardware?
|
|
||
| def add_disks(vim_obj, vmcs, hardware, disks) | ||
| available_units = vim_obj.send(:available_scsi_units, hardware) | ||
| available_scsi_buses = vim_obj.send(:available_scsi_buses, hardware) |
|
Overall it looks fine. Just wonder about the |
|
So unfortunately the answer to the send question is because it was always done that way and I assumed it had something to do with DRb because those methods are being called in the broker. I tested it out without the sends with a broker worker and its just fine so I'll push a commit to remove send from all the calls in this file. |
agrare
force-pushed
the
bz_1445874_vm_reconfigure_scsi_controller
branch
from
e5adccd to
0039fcd
Compare
|
Checked commits agrare/manageiq-providers-vmware@8acb418~...0039fcd with ruby 2.2.6, rubocop 0.47.1, and haml-lint 0.20.0 |
Done |
|
@blomquisg can you merge this one |
|
@agrare @blomquisg this has been requested to be backported to Euwe (which means also Fine). That would be ok? |
|
@simaishi that shouldn't be a problem We'll need ManageIQ/manageiq-gems-pending#148 to go back to those branches as well. |
|
Euwe backport (to manageiq repo) details: |
…troller Lookup SCSI Controller Device Type from Hardware [Depends manageiq-gems-pending/148] (cherry picked from commit 774c485) https://bugzilla.redhat.com/show_bug.cgi?id=1459262
|
Fine backport details: |
Generic Service State Machine method update.
When creating a new SCSI controller we used to always default to
VirtualLsiLogicController, this isn't ideal since some operating systems depend on a certain type of controller.This change will lookup the type from the most recent scsi controller when adding a new one
Depends on: ManageIQ/manageiq-gems-pending#148
Here are some screenshots of adding a new controller when the first controller is of type LSI SAS:
Before:
After:
https://bugzilla.redhat.com/show_bug.cgi?id=1445874