diff options
author | Andrew Gallatin <gallatin@FreeBSD.org> | 2020-12-19 22:04:46 +0000 |
---|---|---|
committer | Andrew Gallatin <gallatin@FreeBSD.org> | 2020-12-19 22:04:46 +0000 |
commit | a034518ac8793059220af22e6ab25f84c5a6ddb8 (patch) | |
tree | cdeacb29f82ad8c1a7d63f8f6da2c77827ec1525 /sys/netinet/in_pcb.h | |
parent | 02bc3865aadf1bfd907bbf555e26719744ebf3c9 (diff) | |
download | src-a034518ac8793059220af22e6ab25f84c5a6ddb8.tar.gz src-a034518ac8793059220af22e6ab25f84c5a6ddb8.zip |
Filter TCP connections to SO_REUSEPORT_LB listen sockets by NUMA domain
In order to efficiently serve web traffic on a NUMA
machine, one must avoid as many NUMA domain crossings as
possible. With SO_REUSEPORT_LB, a number of workers can share a
listen socket. However, even if a worker sets affinity to a core
or set of cores on a NUMA domain, it will receive connections
associated with all NUMA domains in the system. This will lead to
cross-domain traffic when the server writes to the socket or
calls sendfile(), and memory is allocated on the server's local
NUMA node, but transmitted on the NUMA node associated with the
TCP connection. Similarly, when the server reads from the socket,
he will likely be reading memory allocated on the NUMA domain
associated with the TCP connection.
This change provides a new socket ioctl, TCP_REUSPORT_LB_NUMA. A
server can now tell the kernel to filter traffic so that only
incoming connections associated with the desired NUMA domain are
given to the server. (Of course, in the case where there are no
servers sharing the listen socket on some domain, then as a
fallback, traffic will be hashed as normal to all servers sharing
the listen socket regardless of domain). This allows a server to
deal only with traffic that is local to its NUMA domain, and
avoids cross-domain traffic in most cases.
This patch, and a corresponding small patch to nginx to use
TCP_REUSPORT_LB_NUMA allows us to serve 190Gb/s of kTLS encrypted
https media content from dual-socket Xeons with only 13% (as
measured by pcm.x) cross domain traffic on the memory controller.
Reviewed by: jhb, bz (earlier version), bcr (man page)
Tested by: gonzo
Sponsored by: Netfix
Differential Revision: https://reviews.freebsd.org/D21636
Notes
Notes:
svn path=/head/; revision=368819
Diffstat (limited to 'sys/netinet/in_pcb.h')
-rw-r--r-- | sys/netinet/in_pcb.h | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/sys/netinet/in_pcb.h b/sys/netinet/in_pcb.h index 56e2204f9054..080d07cc7218 100644 --- a/sys/netinet/in_pcb.h +++ b/sys/netinet/in_pcb.h @@ -565,7 +565,7 @@ struct inpcblbgroup { struct epoch_context il_epoch_ctx; uint16_t il_lport; /* (c) */ u_char il_vflag; /* (c) */ - u_char il_pad; + u_int8_t il_numa_domain; uint32_t il_pad2; union in_dependaddr il_dependladdr; /* (c) */ #define il_laddr il_dependladdr.id46_addr.ia46_addr4 @@ -852,6 +852,7 @@ int in_pcbinshash(struct inpcb *); int in_pcbinshash_mbuf(struct inpcb *, struct mbuf *); int in_pcbladdr(struct inpcb *, struct in_addr *, struct in_addr *, struct ucred *); +int in_pcblbgroup_numa(struct inpcb *, int arg); struct inpcb * in_pcblookup_local(struct inpcbinfo *, struct in_addr, u_short, int, struct ucred *); |