diff options
author | Ondrej Zajicek <santiago@crfreenet.org> | 2022-12-06 19:40:02 +0100 |
---|---|---|
committer | Ondrej Zajicek <santiago@crfreenet.org> | 2022-12-06 19:51:50 +0100 |
commit | e80156d9363a594ff23524f56d59f0eee160859d (patch) | |
tree | b18539ff6b715e242c292d669792f7729de474d8 /nest/rt-table.c | |
parent | a50d2fa65f3350ee55f5106b87a884d1b98e7761 (diff) |
Nest: Avoid spurious announcements triggered by filtered routes
When filtered routes (enabled by 'import keep filtered' option) are
updated, they trigger announcements by rte_announce(). For regular
channels (e.g. type RA_OPTIMAL or RA_ANY) such announcement is just
ignored, but in case of RA_ACCEPTED (BGP peer with 'secondary' option)
it just reannounces the old (and still valid) best route.
The patch ensures that such no-change is ignored even for these channels.
Diffstat (limited to 'nest/rt-table.c')
-rw-r--r-- | nest/rt-table.c | 15 |
1 files changed, 14 insertions, 1 deletions
diff --git a/nest/rt-table.c b/nest/rt-table.c index cb8c56a6..845b2483 100644 --- a/nest/rt-table.c +++ b/nest/rt-table.c @@ -1112,7 +1112,20 @@ rte_announce(rtable *tab, uint type, net *net, rte *new, rte *old, break; case RA_ACCEPTED: - rt_notify_accepted(c, net, new, old, 0); + /* + * The (new != old) condition is problematic here, as it would break + * the second usage pattern (announcement after bulk change, used in + * rt_next_hop_update_net(), which sends both new and old as NULL). + * + * But recursive next hops do not work with sorted tables anyways, + * such configuration is forbidden in BGP and not supported in + * rt_notify_accepted(). + * + * The condition is needed to eliminate spurious announcements where + * both old and new routes are not valid (so they are NULL). + */ + if (new != old) + rt_notify_accepted(c, net, new, old, 0); break; case RA_MERGED: |