nexthop: fix disappearing addresses and routes after expiration #193
Conversation
This file contains hidden or 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
The boolean logic for determining if a probe should be sent or not is incorrect. We absolutely don't care if the nexthop is a gateway or not. Moreover, request_age > probes does not make any sense. I may not have been in full control of my brain at the time I wrote this. Only try to send probe if a nexthop is either PENDING or STALE. If we already sent the maximum number of probes, mark the nexthop as FAILED so that it is candidate for deletion after one minute. Fixes: 778bd65 ("ip: implement arp aging and refresh") Signed-off-by: Robin Jarry <robin@jarry.cc>
After a fixed period of time (20 minutes), grout will mark non-static nexthops that were resolved earlier (with ARP or NDP) as STALE and will try to probe them to refresh their mac addresses. After 6 probes have been sent (3 unicast using the previously known mac address of the nexthop, and 3 broadcast/multicast) without any reply from the nexthop, it is marked as FAILED. After 1 minute in the FAILED state, it is "freed" using the callback specified in the nexthop_ops per family (at the moment, only IPv4 and IPv6). The current callbacks are rib4_cleanup and rib6_cleanup which are very aggressive. They effectively delete everything related to the nexthop. Including any local address, routes, etc. Replace these callbacks with simpler ones which only delete /32 and /128 routes and decrement the reference counter on the nexthop. If the nexthop still has routes referencing it (for example, it is used as a gateway for a route), scrub its flags and mac address fields so that it can reused later. Fixes: fe3e408 ("route: rework route cleanup routine") Signed-off-by: Robin Jarry <robin@jarry.cc>
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.
After a fixed period of time (20 minutes), grout will mark non-static nexthops that were resolved earlier (with ARP or NDP) as STALE and will try to probe them to refresh their mac addresses.
After 6 probes have been sent (3 unicast using the previously known mac address of the nexthop, and 3 broadcast/multicast) without any reply from the nexthop, it is marked as FAILED.
After 1 minute in the FAILED state, it is "freed" using the callback specified in the nexthop_ops per family (at the moment, only IPv4 and IPv6).
The current callbacks are rib4_cleanup and rib6_cleanup which are very aggressive. They effectively delete everything related to the nexthop. Including any local address, routes, etc.
Replace these callbacks with simpler ones which only delete /32 and /128 routes and decrement the reference counter on the nexthop. If the nexthop still has routes referencing it (for example, it is used as a gateway for a route), scrub its flags and mac address fields so that
it can reused later.
Fixes: fe3e408 ("route: rework route cleanup routine")