Hi Ondrej, did you ever see my v2 patchset? [PATCH v2 1/2] Netlink: Drop ECMP route splitting hacks [PATCH v2 2/2] Netlink: Propagate ECMP nexthop weight to kernel for inet6 routes Since I never heard back on that. --Daniel On Mon, May 23, 2022 at 02:36:49AM +0200, Ondrej Zajicek wrote:
On Sun, May 22, 2022 at 10:28:12PM +0200, dxld@darkboxed.org wrote:
After looking at this more I realise now how bird is splitting multiple ipv6 nexthops into sequences of route updates for the same prefix.
Consequently the code doing this should also be removed. Though it seems it works just fine either way but I get errors for the (now redundant) route removals.
Are we even OK with raising the kernel version required or should I try to preserve support for the old kernel interface?
Hi
The initial IPv6 multipath support required several ugly hacks in Netlink code, which i would be happy to get rid of. When Linux came with IPv6 RTA_MULTIPATH support, the initial versions were somewhat buggy, so we kept using the old API for updates (splitting multipath routes to sequences of route updates), while accepting RTA_MULTIPATH routes in scans / notifications. My notes say that minimal reliable kernel version for IPv6 RTA_MULTIPATH is 4.11.
As a rule of thumb, i target compatibility for ~5 year old systems that are not EOL, for example Debian LTS. Current Debian LTS is Debian 9 (Stretch), which uses kernel 4.9, but it will be EOL in a ~month, while Debian 10 uses kernel 4.19. For kernel itself, ~5 year old LTS release would be 4.14.
So i would be OK with targeting 4.14 / 4.19 as the oldest supported version and switching completely to RTA_MULTIPATH code in Netlink.
-- Elen sila lumenn' omentielvo
Ondrej 'Santiago' Zajicek (email: santiago@crfreenet.org) OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net) "To err is human -- to blame it on a computer is even more so."