Adding support to get safe_reload and check_intf_up_ports #15762
Conversation
|
FYI @abdosi @yejianquan |
| @@ -254,6 +254,8 @@ def reboot(duthost, localhost, reboot_type='cold', delay=10, | |||
| # get 'wait' and 'timeout' from inventory if they are specified, otherwise use current values | |||
| wait = plt_reboot_ctrl.get('wait', wait) | |||
| timeout = plt_reboot_ctrl.get('timeout', timeout) | |||
| safe_reboot = plt_reboot_ctrl.get('safe_reboot', safe_reboot) | |||
There was a problem hiding this comment.
Let's keep using safe_reboot with explicitly parameters, the plt_reboot configuration generally is locally, if we set safe_reboot in configuration, it would cause the differentiates being huge on different platforms.
Any reason we have to set safe_reboot in configuration file?
There was a problem hiding this comment.
Hi @yejianquan, It is used only in places where in the intent is to make sure that the system is fully up before continuing the test. For e.g. in acl, we do a reboot and give a larger wait time for critical process and interfaces to come up and then continue with the tests. The only place that I know of where we don't want to check if critical process is not up and interface is not up is test_reload_config.py because there we want to check almost immediately after reboot, certain conditions. But in all cases we would want to ensure after reboot everything is up. Hence I added that option for the user
There was a problem hiding this comment.
@vperumal Yeah, that's what I supposed,
then let's add safe_reboot=true on the places that call the reboot:
reboot(...., safe_reboot=True)
So that the behaviors keep the same even with different device configurations.
The safe_reboot was an enhanced option.
During the writing of the test cases, it was possible that we don't have safe_reboot option so we didn't use it.
Let's enhance the out-of-date usage of reboot.
|
@vperumal - as discussed in wg, let's create an issue first. |
Description of PR
We had support for time and wait in plt_reboot_dict, Adding support to fetch safe reload and check_intf_up_status. With these parameters we can provide knobs in lab file for which tests vendors want to check critical process status and interface status. This will be particular to their profile and devices, instead of being global to all
For e.g.
plt_reboot_dict:
cold:
timeout: 450
wait: 360
safe_reboot: True
check_intf_up_ports: True
watchdog:
timeout: 300
wait: 360
safe_reboot: True
check_intf_up_ports: True
acl/test_acl.py::TestAclWithReboot:
timeout: 300
wait: 600
safe_reboot: True
check_intf_up_ports: True
platform_tests/test_reload_config.py::test_reload_configuration_checks:
timeout: 300
wait: 60
platform_tests/test_kdump.py::TestKernelPanic::test_kernel_panic:
timeout: 300
wait: 360
safe_reboot: True
check_intf_up_ports: True
Summary:
Fixes # (issue)
Type of change
Back port request
Approach
What is the motivation for this PR?
In some testcases we want to have a provision to check container status and interface status after a reboot.
How did you do it?
How did you verify/test it?
Any platform specific information?
Supported testbed topology if it's a new test case?
Documentation