VNET and BGP route coexistence #3345
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
siqbal1986
changed the title
|
Can you fix coverage? |
prsunny
reviewed
Oct 30, 2024
|
/azpw run |
|
/AzurePipelines run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
8 tasks
prsunny
reviewed
Nov 25, 2024
prsunny
approved these changes
Dec 2, 2024
| @@ -2639,6 +2639,51 @@ bool RouteOrch::removeRoutePost(const RouteBulkContext& ctx) | |||
| return true; | |||
| } | |||
|
|
|||
| bool RouteOrch::hasBgpRoute(const IpPrefix& prefix) | |||
There was a problem hiding this comment.
Please rename to isRouteExists. bgp is not applicable in orchagent.
bradh352
pushed a commit
to bradh352/sonic-swss
that referenced
this pull request
Dec 4, 2024
What I did This PR adds the logic to give Vnet routes precedence over BGP learnt route. This PR also refactors the test_vnet.py to break out the common code into a library. Why I did it Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware.
bradh352
pushed a commit
to bradh352/sonic-swss
that referenced
this pull request
Dec 4, 2024
What I did This PR adds the logic to give Vnet routes precedence over BGP learnt route. This PR also refactors the test_vnet.py to break out the common code into a library. Why I did it Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware.
divyachandralekha
pushed a commit
to divyachandralekha/sonic-swss
that referenced
this pull request
Dec 12, 2024
What I did This PR adds the logic to give Vnet routes precedence over BGP learnt route. This PR also refactors the test_vnet.py to break out the common code into a library. Why I did it Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware.
divyachandralekha
pushed a commit
to divyachandralekha/sonic-swss
that referenced
this pull request
Dec 12, 2024
What I did This PR adds the logic to give Vnet routes precedence over BGP learnt route. This PR also refactors the test_vnet.py to break out the common code into a library. Why I did it Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware.
StormLiangMS
pushed a commit
to sonic-net/sonic-mgmt
that referenced
this pull request
Dec 19, 2024
What is the motivation for this PR? Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware. How I verified it The fix is in PR sonic-net/sonic-swss#3345 These tests cover various scenario in which VNET and BGP routes are added and removed in different order How did you do it? How did you verify/test it? Any platform specific information? Cisco-8000 and mlnx. Supported testbed topology if it's a new test case? Documentation There are 5 differnet scenarios which are checked. Each scenario is tested with with the following options. Encap types [v4_inv4, v6_inV4] Monitor Type [BFD, Custom] Init NH state( BFD/monitor sessions for nexthops are initially up or not.) test_vnet_route_after_bgp ADD BGP ROUTE on TOR Add VNET route Configure monitor (BFD or custom) with nexthop state (UP) Test with traffic Remove VNET route Remove BGP route test_vnet_route_before_bgp_after_ep_up Add VNET route Configure monitor (BFD or custom) with nexthop state (UP) Add BGP ROUTE on TOR Test with traffic Remove VNET ROUTE Remove BGP route test_vnet_route_bgp_removal_before_ep ADD BGP ROUTE on TOR Add VNET route Remove BGP route Configure monitor (BFD or custom) with nexthop state (UP) Test with traffic Remove VNET route test_vnet_route_after_bgp_with_early_bgp_removal Add VNET route Add BGP ROUTE on TOR Configure monitor (BFD or custom) with nexthop state (UP) Test with traffic Remove BGP route Test with traffic Remove VNET route test_vnet_route_after_bgp_multi_flap ADD BGP ROUTE on TOR Add VNET route Configure monitor (BFD or custom) with nexthop state (UP) Test with traffic flap the bfd/monitor sessions. Test with traffic Remove VNET route Remove BGP route
stepanblyschak
pushed a commit
to stepanblyschak/sonic-swss
that referenced
this pull request
Jan 27, 2025
What I did This PR adds the logic to give Vnet routes precedence over BGP learnt route. This PR also refactors the test_vnet.py to break out the common code into a library. Why I did it Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware.
shiraez
pushed a commit
to Marvell-switching/sonic-swss
that referenced
this pull request
Feb 17, 2025
What I did This PR adds the logic to give Vnet routes precedence over BGP learnt route. This PR also refactors the test_vnet.py to break out the common code into a library. Why I did it Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What I did
This PR adds the logic to give Vnet routes precedence over BGP learnt route.
This PR also refactors the test_vnet.py to break out the common code into a library.
Why I did it
Currently if a route is learnt via BGP and added into the hardware, adding a VNET route results in failure. In addition due to a bug in VnetOrch, we start advertising the failed route, This prompts the BGP to remove the learnt route in favor of the local route. Since the Vnet Orch doesn't retry adding the Vnet route, This results in no route being present in the Hardware.
How I verified it
Added 2 tests to verify the learnt route is removed when a Vnet route is added.
One test covers the scnerio where BFD sessions are brought up before the Vnet route is added and the 2nd covers when the Vnet route is added before the BFD sessions are brought up.
Details if related
Tests covering this chage.
sonic-net/sonic-mgmt#15710