pimd: Ensure upstream points at the correct rpf #14707
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
In the scenario on an intermediate router where a *,G join has been received and a S,G stream is being sent through that router on the *,G stream, there exists a situation when the *,G in has been pruned but the stream is still being received on on incoming interface towards the RP for the *,G. In this situation PIM will see the S,G stream initially as a NOCACHE from the dataplane, PIM will then do a RPF for the S and notice that it is supposed to be coming in on adifferent interface. In this case PIM the original PIM code would create a blackhole mroute towards the RPF of the *,G( the interface the stream is being received on ). The original reason for this is that if there is a scenario where this particular S1,G stream is sending at basically line rate, and there also happens to be a different S2,G stream that is sending at a very low rate. With certain dataplanes there is no way to really rate limit the S1 -vs- S2 stream and the S1 stream completely overwhelms the S2 stream for sending up to the control plane for proper pim handling. The problem then becomes that FRR never properly responds to the situation where the *,G is rereceived and the S,G stream switches back over to the SPT for itself and FRR ends up with a dead mroute that stops everything from working properly. This code change, installs the blackhole mroute with the RPF towards the RP for the G and then resets the RPF to the correct RPF for the Stream but does not modify the mroute. When the *,G is rereceived and we attempt to transition to the S,G stream this now works. As a note: Both David L and myself do not necessarily believe we fully understand the problem yet. What this does do is fix all the inconsistent CI issues we are seeing in the topotests at this time. Internally I am seeing other test failures in PIM that I don't fully understand and we suspect that there are other problems in the state machine. We plan to revisit this problem as we are able to debug the issue better. In the meantime both David and Myself agree that this gets the CI working again and Streams end up in the right state. Signed-off-by: Donald Sharp <[email protected]>
|
@Mergifyio backport dev/9.1 |
✅ Backports have been created
|
donaldsharp
added a commit
that referenced
this pull request
Nov 2, 2023
pimd: Ensure upstream points at the correct rpf (backport #14707)
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.
In the scenario on an intermediate router where a *,G join has been received and a S,G stream is being sent through that router on the *,G stream, there exists a situation when the *,G in has been pruned but the stream is still being received on on incoming interface towards
the RP for the *,G. In this situation PIM will see the S,G stream
initially as a NOCACHE from the dataplane, PIM will then do a RPF
for the S and notice that it is supposed to be coming in on adifferent
interface. In this case PIM the original PIM code would create
a blackhole mroute towards the RPF of the *,G( the interface the
stream is being received on ). The original reason for this is that
if there is a scenario where this particular S1,G stream is sending
at basically line rate, and there also happens to be a different
S2,G stream that is sending at a very low rate. With certain
dataplanes there is no way to really rate limit the S1 -vs- S2
stream and the S1 stream completely overwhelms the S2 stream
for sending up to the control plane for proper pim handling.
The problem then becomes that FRR never properly responds
to the situation where the *,G is rereceived and the S,G
stream switches back over to the SPT for itself and FRR ends
up with a dead mroute that stops everything from working properly.
This code change, installs the blackhole mroute with the RPF towards the RP for the G and then resets the RPF to the correct RPF for the Stream but does not modify the mroute. When the *,G is rereceived and we attempt to transition to the S,G stream this now works.
As a note: Both David L and myself do not necessarily believe we fully understand the problem yet. What this does do is fix all the inconsistent CI issues we are seeing in the topotests at this time. Internally I am seeing other test failures in PIM that I don't fully understand and we suspect that there are other problems in the state machine. We plan to revisit this problem as we are able to debug the issue better. In the meantime both David and Myself agree that this gets the CI working again and Streams end up in the right state.