-
Notifications
You must be signed in to change notification settings - Fork 657
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[VRF]Adding CLI checks to ensure Vrf is valid in interface bind and static route commands #2333
Conversation
…tatic route commands
@venkatmahalingam @prsunny Can you please review? |
@@ -998,6 +1012,8 @@ def cli_sroute_to_config(ctx, command_str, strict_nh = True): | |||
if 'prefix' in prefix_str and 'vrf' in prefix_str: | |||
# prefix_str: ['prefix', 'vrf', Vrf-name, ip] | |||
vrf_name = prefix_str[2] | |||
if not is_vrf_exists(config_db, vrf_name): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Do we have check to ensure vrf is removed only after all interface bindings are removed? Otherwise this is only a check in the add-path.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is only a check in the add path.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In fact the current vrf removal logic removes all these bindings implicitly using the function del_interface_bind_to_vrf
return True | ||
elif vrf_name == "mgmt": | ||
entry = config_db.get_entry("MGMT_VRF_CONFIG", "vrf_global") | ||
if entry and entry.get("mgmtVrfEnabled") == "true": |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
How do we prevent from using non-existing VRF in the INTERFACE table during config load? this seems to address only in the Click CLI.
If we can do the YANG level validation for all north bound interfaces (CLick, config load..etc), it would be an one time check in the back-end rather than doing it on all front-end (e.g this check in Click), what is your thought?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @venkatmahalingam
Your comment is applicable for any table in config_db.json.
This PR is focussed on only the Click CLI. I believe we have a separate discussion on using yang for validation which is the work in progress and would take sometime.
For now It is expected of the end user to use Click CLI to make any incremental configuration changes.
Let me know if the changes look good.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Any update we do with config-db should be YANG validated is what the expectation but the infra is not ready yet, we can merge this code for now.
…tatic route commands (#2333) - What I did Added CLI checks to verify if Vrf is created and valid when used in interface bind and static route commands - How I did it Check for presence of Vrf in VRF table or in case if it is mgmt vrf, verify if mgmt vrf is enabled/ - How to verify it Added UT to verify it.
Update sonic-utilities submodule pointer to include the following: * [route_check]: Ignore standalone tunnel routes (sonic-net#2325) ([sonic-net#2346](sonic-net/sonic-utilities#2346)) * [VRF]Adding CLI checks to ensure Vrf is valid in interface bind and static route commands ([sonic-net#2333](sonic-net/sonic-utilities#2333)) * Subinterface vrf bind issue fix ([sonic-net#2211](sonic-net/sonic-utilities#2211)) * [decode-syseeprom] Fix setting use_db based on support_eeprom_db ([sonic-net#2270](sonic-net/sonic-utilities#2270)) * Fix vrf UT failed issue ([sonic-net#2309](sonic-net/sonic-utilities#2309)) Signed-off-by: dprital <drorp@nvidia.com>
Update sonic-utilities submodule pointer to include the following: * [route_check]: Ignore standalone tunnel routes (#2325) ([#2346](sonic-net/sonic-utilities#2346)) * [VRF]Adding CLI checks to ensure Vrf is valid in interface bind and static route commands ([#2333](sonic-net/sonic-utilities#2333)) * Subinterface vrf bind issue fix ([#2211](sonic-net/sonic-utilities#2211)) * [decode-syseeprom] Fix setting use_db based on support_eeprom_db ([#2270](sonic-net/sonic-utilities#2270)) * Fix vrf UT failed issue ([#2309](sonic-net/sonic-utilities#2309)) Signed-off-by: dprital <drorp@nvidia.com> Signed-off-by: dprital <drorp@nvidia.com>
…tatic route commands (sonic-net#2333) - What I did Added CLI checks to verify if Vrf is created and valid when used in interface bind and static route commands - How I did it Check for presence of Vrf in VRF table or in case if it is mgmt vrf, verify if mgmt vrf is enabled/ - How to verify it Added UT to verify it.
What I did
Added CLI checks to verify if Vrf is created and valid when used in interface bind and static route commands
How I did it
Check for presence of Vrf in VRF table or in case if it is mgmt vrf, verify if mgmt vrf is enabled/
How to verify it
Added UT to verify it.
Previous command output (if the output of a command-line utility has changed)
New command output (if the output of a command-line utility has changed)