mptcp: improve mptcp-level window tracking
Subsystems (3)
Paolo Abeni
maintainerv2 cover mptcp: improve mptcp-level window tracking
I've been chasing bad/unstable performances with multiple subflows on very high speed links.
It looks like the root cause is due to the current mptcp-level congestion window handling. There are apparently a few different sub-issues:
- the rcv_wnd is not effectively shared on the tx side, as each subflow takes in account only the value received by the underlaying TCP connection. This is addressed in patch 1/4
- The mptcp-level offered wnd right edge is currently allowed to shrink. Reading section 3.3.4.:
"""
The receive window is relative to the DATA_ACK. As in TCP, a receiver MUST NOT shrink the right edge of the receive window (i.e., DATA_ACK + receive window). The receiver will use the data sequence number to tell if a packet should be accepted at the connection level.
"""
I read the above as we need to reflect window right-edge tracking on the wire, see patch 3/4.
- The offered window right edge tracking can happen concurrently on multiple subflows, but there is no mutex protection. We need an additional atomic operation - still patch 3/4
This series additionally bump a few new MIBs to track all the above (ensure/observe that the suspected races actually take place).
I could not access again the host where the issue was su much noticeable, still in the current setup the tput changes from [6-18] Gbps to 19Gbps very stable.
v1 -> v2:
- pass only the TCP header to tcp_options_write (Mat)
- fix build issues on some 32 bit arches (intel bot)
RFC -> v1:
- added patch 3/5 to address Mat's comment, and rebased the following on top of it - I hope Eric may tolerate that, it's more an hope than guess ;)
Paolo Abeni (5):
mptcp: really share subflow snd_wnd
mptcp: add mib for xmit window sharing
tcp: allow MPTCP to update the announced window.
mptcp: never shrink offered window
mptcp: add more offered MIBs counter.
include/net/mptcp.h | 2 +- net/ipv4/tcp_output.c | 14 ++++++----- net/mptcp/mib.c | 4 +++ net/mptcp/mib.h | 6 +++++ net/mptcp/options.c | 58 +++++++++++++++++++++++++++++++++++++------ net/mptcp/protocol.c | 32 +++++++++++++++--------- net/mptcp/protocol.h | 2 +- 7 files changed, 90 insertions(+), 28 deletions(-)
Signature · 1 line
2.35.1
Paolo Abeni
maintainerPATCH v2 1/5 mptcp: really share subflow snd_wnd
As per RFC, mptcp subflows use a "shared" snd_wnd: the effective window is the maximum among the current values received on all subflows. Without such feature a data transfer using multiple subflows could block.
Window sharing is currently implemented in the RX side:__tcp_select_window uses the mptcp-level receive buffer to compute the announced window.
That is not enough: the TCP stack will stick to the window size received on the given subflow; we need to propagate the msk window value on each subflow at xmit time.
Change the packet scheduler to ignore the subflow level window and use instead the msk level one
--- net/mptcp/protocol.c | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-)
net/mptcp/protocol.c +15 -9
@@ -1141,19 +1141,20 @@ struct mptcp_sendmsg_info {bool data_lock_held; }; -static int mptcp_check_allowed_size (struct mptcp_sock *msk,u64 data_seq, -int avail_size) +static int mptcp_check_allowed_size (const struct mptcp_sock *msk,struct sock *ssk, +u64 data_seq,int avail_size) {u64 window_end =mptcp_wnd_end (msk); +u64 mptcp_snd_wnd;if (__mptcp_check_fallback (msk))return avail_size; -if (!before64 (data_seq + avail_size, window_end)) { -u64 allowed_size = window_end - data_seq; + mptcp_snd_wnd = window_end - data_seq; + avail_size = min_t(unsigned int, mptcp_snd_wnd, avail_size); - return min_t(unsigned int, allowed_size, avail_size); - } +if (unlikely (tcp_sk (ssk)->snd_wnd < mptcp_snd_wnd)) +tcp_sk (ssk)->snd_wnd =min_t (u64,U32_MAX , mptcp_snd_wnd);return avail_size; }
@@ -1305,7 +1306,7 @@ static int mptcp_sendmsg_frag(struct sock *sk, struct sock *ssk, }/* Zero window and all data acked? Probe. */ - copy =mptcp_check_allowed_size (msk, data_seq, copy); + copy =mptcp_check_allowed_size (msk, ssk, data_seq, copy);if (copy ==0 ) {u64 snd_una =READ_ONCE (msk->snd_una);
@@ -1498,11 +1499,16 @@ static struct sock *mptcp_subflow_get_send(struct mptcp_sock *msk)* to check that subflow has a non empty cwin. */ ssk = send_info[SSK_MODE_ACTIVE ].ssk; -if (!ssk || !sk_stream_memory_free (ssk) || !tcp_sk (ssk)->snd_wnd) +if (!ssk || !sk_stream_memory_free (ssk))return NULL ; - burst =min_t (int,MPTCP_SEND_BURST_SIZE ,tcp_sk (ssk)->snd_wnd); + burst =min_t (int,MPTCP_SEND_BURST_SIZE ,mptcp_wnd_end (msk) - msk->snd_nxt); wmem =READ_ONCE (ssk->sk_wmem_queued); +if (!burst) { + msk->last_snd =NULL ; +return ssk; + } + subflow =mptcp_subflow_ctx (ssk); subflow->avg_pacing_rate = div_u64((u64)subflow->avg_pacing_rate * wmem + READ_ONCE(ssk->sk_pacing_rate) * burst,
Signature · 1 line
2.35.1
Paolo Abeni
maintainerPATCH v2 2/5 mptcp: add mib for xmit window sharing
In reply to Paolo AbeniBump a counter for counter when snd_wnd is shared among subflow, for observability's sake.
--- net/mptcp/mib.c | 1 + net/mptcp/mib.h | 1 + net/mptcp/protocol.c | 4 +++- 3 files changed, 5 insertions(+), 1 deletion(-)
net/mptcp/mib.c +1 -0
@@ -56,6 +56,7 @@ static const struct snmp_mib mptcp_snmp_list[] = {SNMP_MIB_ITEM ("RcvPruned" ,MPTCP_MIB_RCVPRUNED ),SNMP_MIB_ITEM ("SubflowStale" ,MPTCP_MIB_SUBFLOWSTALE ),SNMP_MIB_ITEM ("SubflowRecover" ,MPTCP_MIB_SUBFLOWRECOVER ), +SNMP_MIB_ITEM ("SndWndShared" ,MPTCP_MIB_SNDWNDSHARED ), SNMP_MIB_SENTINEL };
net/mptcp/mib.h +1 -0
@@ -49,6 +49,7 @@ enum linux_mptcp_mib_field {MPTCP_MIB_RCVPRUNED ,/* Incoming packet dropped due to memory limit */ MPTCP_MIB_SUBFLOWSTALE ,/* Subflows entered 'stale' status */ MPTCP_MIB_SUBFLOWRECOVER ,/* Subflows returned to active status after being stale */ +MPTCP_MIB_SNDWNDSHARED ,/* Subflow snd wnd is overridden by msk's one */ __MPTCP_MIB_MAX };
net/mptcp/protocol.c +3 -1
@@ -1153,8 +1153,10 @@ static int mptcp_check_allowed_size(const struct mptcp_sock *msk, struct sock *s mptcp_snd_wnd = window_end - data_seq; avail_size = min_t(unsigned int, mptcp_snd_wnd, avail_size); -if (unlikely (tcp_sk (ssk)->snd_wnd < mptcp_snd_wnd)) +if (unlikely (tcp_sk (ssk)->snd_wnd < mptcp_snd_wnd)) {tcp_sk (ssk)->snd_wnd =min_t (u64,U32_MAX , mptcp_snd_wnd); +MPTCP_INC_STATS (sock_net (ssk),MPTCP_MIB_SNDWNDSHARED ); + }return avail_size; }
Signature · 1 line
2.35.1
Paolo Abeni
maintainerPATCH v2 3/5 tcp: allow MPTCP to update the announced window.
In reply to Paolo AbeniThe MPTCP RFC requires that the MPTCP-level receive window's right edge never moves backward. Currently the MPTCP code enforces such constraint while tracking the right edge, but it does not reflects it on the wire, as MPTCP lacks a suitable hook to update accordingly the TCP header.
This change modifies the existing mptcp_write_options() hook, providing the current packet's TCP header to the MPTCP protocol, so that the next patch could implement the above mentioned constraint.
No functional changes intended.
---
v1 -> v2:
- Use a single hdr/opt argument (Mat)
--- include/net/mptcp.h | 2 +- net/ipv4/tcp_output.c | 14 ++++++++------ net/mptcp/options.c | 2 +- 3 files changed, 10 insertions(+), 8 deletions(-)
include/net/mptcp.h +1 -1
@@ -125,7 +125,7 @@ bool mptcp_established_options(struct sock *sk, struct sk_buff *skb, struct mptcp_out_options *opts);bool mptcp_incoming_options (struct sock *sk,struct sk_buff *skb); -void mptcp_write_options (__be32 *ptr,const struct tcp_sock *tp, +void mptcp_write_options (struct tcphdr *th,__be32 *ptr,struct tcp_sock *tp,struct mptcp_out_options *opts);void mptcp_diag_fill_info (struct mptcp_sock *msk,struct mptcp_info *info);
net/ipv4/tcp_output.c +8 -6
@@ -444,12 +444,13 @@ struct tcp_out_options {struct mptcp_out_options mptcp; }; -static void mptcp_options_write (__be32 *ptr,const struct tcp_sock *tp, +static void mptcp_options_write (struct tcphdr *th,__be32 *ptr, +struct tcp_sock *tp,struct tcp_out_options *opts) {#if IS_ENABLED (CONFIG_MPTCP )if (unlikely (OPTION_MPTCP & opts->options)) -mptcp_write_options (ptr, tp, &opts->mptcp); +mptcp_write_options (th, ptr, tp, &opts->mptcp);#endif }
@@ -605,9 +606,10 @@ static void bpf_skops_write_hdr_opt(struct sock *sk, struct sk_buff *skb,* At least SACK_PERM as the first option is known to lead to a disaster * (but it may well be that other scenarios fail similarly). */ -static void tcp_options_write (__be32 *ptr,struct tcp_sock *tp, +static void tcp_options_write (struct tcphdr *th,struct tcp_sock *tp,struct tcp_out_options *opts) { +__be32 *ptr = (__be32 *)(th +1 );u16 options = opts->options;/* mungable copy */ if (unlikely (OPTION_MD5 & options)) {
@@ -701,7 +703,7 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp,smc_options_write (ptr, &options); -mptcp_options_write (ptr, tp, opts); +mptcp_options_write (th, ptr, tp, opts); } static void smc_set_option(const struct tcp_sock *tp,
@@ -1354,7 +1356,7 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, th->window =htons (min (tp->rcv_wnd,65535U )); } -tcp_options_write ((__be32 *)(th +1 ), tp, &opts); +tcp_options_write (th, tp, &opts);#ifdef CONFIG_TCP_MD5SIG /* Calculate the MD5 hash, as we have all we need now */
@@ -3590,7 +3592,7 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst,/* RFC1323: The window in SYN & SYN/ACK segments is never scaled. */ th->window =htons (min (req->rsk_rcv_wnd,65535U )); -tcp_options_write ((__be32 *)(th +1 ),NULL , &opts); +tcp_options_write (th,NULL , &opts); th->doff = (tcp_header_size >>2 );__TCP_INC_STATS (sock_net (sk),TCP_MIB_OUTSEGS );
net/mptcp/options.c +1 -1
@@ -1265,7 +1265,7 @@ static u16 mptcp_make_csum(const struct mptcp_ext *mpext) ~csum_unfold(mpext->csum)); } -void mptcp_write_options (__be32 *ptr,const struct tcp_sock *tp, +void mptcp_write_options (struct tcphdr *th,__be32 *ptr,struct tcp_sock *tp,struct mptcp_out_options *opts) {const struct sock *ssk = (const struct sock *)tp;
Signature · 1 line
2.35.1
Paolo Abeni
maintainerPATCH v2 4/5 mptcp: never shrink offered window
In reply to Paolo AbeniAs per RFC, the offered MPTCP-level window should never shrink. While we currently track the right edge, we don't enforce the above constraint on the wire.
Additionally, concurrent xmit on different subflows can end-up in erroneous right edge update.
Address the above explicitly updating the announced window and protecting the update with an additional atomic operation (sic)
---
v1 -> v2:
- convert rcv_wnd_sent to atomic64: cmpxchg64 is not supported by every arch, while atomic64_t has no such issue
RFC -> v1:
- rebased on previous patch
--- net/mptcp/options.c | 52 ++++++++++++++++++++++++++++++++++++++------ net/mptcp/protocol.c | 8 +++---- net/mptcp/protocol.h | 2 +- 3 files changed, 50 insertions(+), 12 deletions(-)
net/mptcp/options.c +45 -7
@@ -1224,20 +1224,58 @@ bool mptcp_incoming_options(struct sock *sk, struct sk_buff *skb)return true; } -static void mptcp_set_rwin (const struct tcp_sock *tp) +static void mptcp_set_rwin (struct tcp_sock *tp,struct tcphdr *th) {const struct sock *ssk = (const struct sock *)tp; -const struct mptcp_subflow_context *subflow; +struct mptcp_subflow_context *subflow; +u64 ack_seq, rcv_wnd_old, rcv_wnd_new;struct mptcp_sock *msk; -u64 ack_seq; +u32 new_win; +u64 win; subflow =mptcp_subflow_ctx (ssk); msk =mptcp_sk (subflow->conn); - ack_seq =READ_ONCE (msk->ack_seq) + tp->rcv_wnd; + ack_seq =READ_ONCE (msk->ack_seq); + rcv_wnd_new = ack_seq + tp->rcv_wnd; + + rcv_wnd_old =atomic64_read (&msk->rcv_wnd_sent); +if (after64 (rcv_wnd_new, rcv_wnd_old)) { +u64 rcv_wnd; + +for (;;) { + rcv_wnd =atomic64_cmpxchg (&msk->rcv_wnd_sent, rcv_wnd_old, rcv_wnd_new); + +if (rcv_wnd == rcv_wnd_old) +break ; +if (before64 (rcv_wnd_new, rcv_wnd)) + goto raise_win; + rcv_wnd_old = rcv_wnd; + }; +return ; + } + +if (rcv_wnd_new != rcv_wnd_old) { +raise_win: + win = rcv_wnd_old - ack_seq; + tp->rcv_wnd =min_t (u64, win,U32_MAX ); + new_win = tp->rcv_wnd; -if (after64 (ack_seq,READ_ONCE (msk->rcv_wnd_sent))) -WRITE_ONCE (msk->rcv_wnd_sent, ack_seq); +/* Make sure we do not exceed the maximum possible +* scaled window. +*/ +if (unlikely (th->syn)) + new_win =min (new_win,65535U ) << tp->rx_opt.rcv_wscale; +if (!tp->rx_opt.rcv_wscale && +sock_net (ssk)->ipv4.sysctl_tcp_workaround_signed_windows) + new_win =min (new_win,MAX_TCP_WINDOW ); +else + new_win =min (new_win, (65535U << tp->rx_opt.rcv_wscale)); + +/* RFC1323 scaling applied */ + new_win >>= tp->rx_opt.rcv_wscale; + th->window =htons (new_win); + } }u16 __mptcp_make_csum (u64 data_seq,u32 subflow_seq,u16 data_len,__wsum sum)
@@ -1554,7 +1592,7 @@ void mptcp_write_options(struct tcphdr *th, __be32 *ptr, struct tcp_sock *tp, }if (tp) -mptcp_set_rwin (tp); +mptcp_set_rwin (tp, th); }__be32 mptcp_get_reset_option (const struct sk_buff *skb)
net/mptcp/protocol.c +4 -4
@@ -216,7 +216,7 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk, struct sk_buff *skb) seq =MPTCP_SKB_CB (skb)->map_seq; end_seq =MPTCP_SKB_CB (skb)->end_seq; - max_seq =READ_ONCE (msk->rcv_wnd_sent); + max_seq =atomic64_read (&msk->rcv_wnd_sent);pr_debug ("msk=%p seq=%llx limit=%llx empty=%d" , msk, seq, max_seq,RB_EMPTY_ROOT (&msk->out_of_order_queue));
@@ -225,7 +225,7 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk, struct sk_buff *skb)mptcp_drop (sk, skb);pr_debug ("oow by %lld, rcv_wnd_sent %llu\n" , (unsigned long long )end_seq - (unsigned long )max_seq, - (unsigned long long )msk->rcv_wnd_sent); + (unsigned long long )atomic64_read (&msk->rcv_wnd_sent));MPTCP_INC_STATS (sock_net (sk),MPTCP_MIB_NODSSWINDOW );return ; }
@@ -3008,7 +3008,7 @@ struct sock *mptcp_sk_clone(const struct sock *sk,mptcp_crypto_key_sha (msk->remote_key,NULL , &ack_seq); ack_seq++;WRITE_ONCE (msk->ack_seq, ack_seq); -WRITE_ONCE (msk->rcv_wnd_sent, ack_seq); +atomic64_set (&msk->rcv_wnd_sent, ack_seq); }#if !IS_ENABLED (CONFIG_KASAN )
@@ -3303,9 +3303,9 @@ void mptcp_finish_connect(struct sock *ssk)WRITE_ONCE (msk->write_seq, subflow->idsn +1 );WRITE_ONCE (msk->snd_nxt, msk->write_seq);WRITE_ONCE (msk->ack_seq, ack_seq); -WRITE_ONCE (msk->rcv_wnd_sent, ack_seq);WRITE_ONCE (msk->can_ack,1 );WRITE_ONCE (msk->snd_una, msk->write_seq); +atomic64_set (&msk->rcv_wnd_sent, ack_seq);mptcp_pm_new_connection (msk, ssk,0 );
net/mptcp/protocol.h +1 -1
@@ -257,7 +257,7 @@ struct mptcp_sock {u64 write_seq;u64 snd_nxt;u64 ack_seq; -u64 rcv_wnd_sent; +atomic64_t rcv_wnd_sent;u64 rcv_data_fin_seq;int rmem_fwd_alloc;struct sock *last_snd;
Signature · 1 line
2.35.1
Paolo Abeni
maintainerPATCH v2 5/5 mptcp: add more offered MIBs counter.
In reply to Paolo AbeniTrack the exceptional handling of MPTCP-level offered window with a few more counters for observability.
--- net/mptcp/mib.c | 3 +++ net/mptcp/mib.h | 5 +++++ net/mptcp/options.c | 6 +++++- 3 files changed, 13 insertions(+), 1 deletion(-)
net/mptcp/mib.c +3 -0
@@ -57,6 +57,9 @@ static const struct snmp_mib mptcp_snmp_list[] = {SNMP_MIB_ITEM ("SubflowStale" ,MPTCP_MIB_SUBFLOWSTALE ),SNMP_MIB_ITEM ("SubflowRecover" ,MPTCP_MIB_SUBFLOWRECOVER ),SNMP_MIB_ITEM ("SndWndShared" ,MPTCP_MIB_SNDWNDSHARED ), +SNMP_MIB_ITEM ("RcvWndShared" ,MPTCP_MIB_RCVWNDSHARED ), +SNMP_MIB_ITEM ("RcvWndConflictUpdate" ,MPTCP_MIB_RCVWNDCONFLICTUPDATE ), +SNMP_MIB_ITEM ("RcvWndConflict" ,MPTCP_MIB_RCVWNDCONFLICT ), SNMP_MIB_SENTINEL };
net/mptcp/mib.h +5 -0
@@ -50,6 +50,11 @@ enum linux_mptcp_mib_field {MPTCP_MIB_SUBFLOWSTALE ,/* Subflows entered 'stale' status */ MPTCP_MIB_SUBFLOWRECOVER ,/* Subflows returned to active status after being stale */ MPTCP_MIB_SNDWNDSHARED ,/* Subflow snd wnd is overridden by msk's one */ +MPTCP_MIB_RCVWNDSHARED ,/* Subflow rcv wnd is overridden by msk's one */ +MPTCP_MIB_RCVWNDCONFLICTUPDATE ,/* subflow rcv wnd is overridden by msk's one due to +* conflict with another subflow while updating msk rcv wnd +*/ +MPTCP_MIB_RCVWNDCONFLICT ,/* Conflict with while updating msk rcv wnd */ __MPTCP_MIB_MAX };
net/mptcp/options.c +5 -1
@@ -1248,8 +1248,11 @@ static void mptcp_set_rwin(struct tcp_sock *tp, struct tcphdr *th)if (rcv_wnd == rcv_wnd_old)break ; -if (before64 (rcv_wnd_new, rcv_wnd)) +if (before64 (rcv_wnd_new, rcv_wnd)) { +MPTCP_INC_STATS (sock_net (ssk),MPTCP_MIB_RCVWNDCONFLICTUPDATE ); goto raise_win; + } +MPTCP_INC_STATS (sock_net (ssk),MPTCP_MIB_RCVWNDCONFLICT ); rcv_wnd_old = rcv_wnd; };return ;
@@ -1275,6 +1278,7 @@ static void mptcp_set_rwin(struct tcp_sock *tp, struct tcphdr *th)/* RFC1323 scaling applied */ new_win >>= tp->rx_opt.rcv_wscale; th->window =htons (new_win); +MPTCP_INC_STATS (sock_net (ssk),MPTCP_MIB_RCVWNDSHARED ); } }
Signature · 1 line
2.35.1
Re: mptcp: add more offered MIBs counter.: Tests Results
Hi Paolo,
Thank you for your modifications, that's great!
Our CI did some validations and here is its report:
- KVM Validation: normal:
- Success! ✅:
- Task: https://cirrus-ci.com/task/5291819050205184
- Summary: https://api.cirrus-ci.com/v1/artifact/task/5291819050205184/summary/summary.txt
- KVM Validation: debug:
- Unstable: 2 failed test(s): selftest_diag selftest_mptcp_join - Critical: KMemLeak ❌:
- Task: https://cirrus-ci.com/task/6417718957047808
- Summary: https://api.cirrus-ci.com/v1/artifact/task/6417718957047808/summary/summary.txt
Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/d2204656a1fb
If there are some issues, you can reproduce them using the same environment as the one used by the CI thanks to a docker image, e.g.:
$ cd [kernel source code]
$ docker run -v "${PWD}:${PWD}:rw" -w "${PWD}" --privileged --rm -it \
--pull always mptcp/mptcp-upstream-virtme-docker:latest \
auto-debugFor more details:
https://github.com/multipath-tcp/mptcp-upstream-virtme-docker
Please note that despite all the efforts that have been already done to have a stable tests suite when executed on a public CI like here, it is possible some reported issues are not due to your modifications. Still, do not hesitate to help us improve that ;-)
Cheers,
MPTCP GH Action bot
Bot operated by Matthieu Baerts (Tessares)
Quoted 32 lines · Paolo Abeni
I've been chasing bad/unstable performances with multiple subflows on very high speed links.
It looks like the root cause is due to the current mptcp-level congestion window handling. There are apparently a few different sub-issues:
- the
rcv_wndis not effectively shared on the tx side, as each subflow takes in account only the value received by the underlaying TCP connection. This is addressed in patch 1/4- The mptcp-level offered wnd right edge is currently allowed to shrink. Reading section 3.3.4.:
"""
The receive window is relative to theDATA_ACK. As in TCP, a receiver MUST NOT shrink the right edge of the receive window (i.e.,DATA_ACK+ receive window). The receiver will use the data sequence number to tell if a packet should be accepted at the connection level.
"""I read the above as we need to reflect window right-edge tracking on the wire, see patch 3/4.
- The offered window right edge tracking can happen concurrently on multiple subflows, but there is no mutex protection. We need an additional atomic operation - still patch 3/4
This series additionally bump a few new MIBs to track all the above (ensure/observe that the suspected races actually take place).
Quoted 7 lines
I could not access again the host where the issue was su much noticeable, still in the current setup the tput changes from [6-18] Gbps to 19Gbps very stable.
v1 -> v2:
- pass only the TCP header totcp_options_write(Mat)
- fix build issues on some 32 bit arches (intel bot)
v2 looks good for the export branch, thanks Paolo.
Quoted 13 lines
RFC -> v1:
- added patch 3/5 to address Mat's comment, and rebased the following on top of it - I hope Eric may tolerate that, it's more an hope than guess ;)Paolo Abeni (5):
mptcp: really share subflowsnd_wnd
mptcp: add mib for xmit window sharing
tcp: allow MPTCP to update the announced window.
mptcp: never shrink offered window
mptcp: add more offered MIBs counter.
Quoted 14 lines
include/net/mptcp.h | 2 +- net/ipv4/tcp_output.c | 14 ++++++----- net/mptcp/mib.c | 4 +++ net/mptcp/mib.h | 6 +++++ net/mptcp/options.c | 58 +++++++++++++++++++++++++++++++++++++------ net/mptcp/protocol.c | 32 +++++++++++++++--------- net/mptcp/protocol.h | 2 +- 7 files changed, 90 insertions(+), 28 deletions(-)Signature · 1 line
2.35.1
--
Mat Martineau
Intel
Hi Paolo, Mat,
Quoted 29 lines · Paolo Abeni
I've been chasing bad/unstable performances with multiple subflows on very high speed links.
It looks like the root cause is due to the current mptcp-level congestion window handling. There are apparently a few different sub-issues:
- the
rcv_wndis not effectively shared on the tx side, as each subflow takes in account only the value received by the underlaying TCP connection. This is addressed in patch 1/4- The mptcp-level offered wnd right edge is currently allowed to shrink. Reading section 3.3.4.:
"""
The receive window is relative to theDATA_ACK. As in TCP, a receiver MUST NOT shrink the right edge of the receive window (i.e.,DATA_ACK+ receive window). The receiver will use the data sequence number to tell if a packet should be accepted at the connection level.
"""I read the above as we need to reflect window right-edge tracking on the wire, see patch 3/4.
- The offered window right edge tracking can happen concurrently on multiple subflows, but there is no mutex protection. We need an additional atomic operation - still patch 3/4
Quoted 6 lines
This series additionally bump a few new MIBs to track all the above (ensure/observe that the suspected races actually take place).
I could not access again the host where the issue was su much noticeable, still in the current setup the tput changes from [6-18] Gbps to 19Gbps very stable.
Thank you for the patches and reviews!
Now in our tree (feat. for net-next) with Mat's RvB tag:
New patches for t/upstream:
- 61ed3e818378: mptcp: really share subflow snd_wnd
- d5b00c55441f: mptcp: add mib for xmit window sharing
- 0b814d52c6bb: tcp: allow MPTCP to update the announced window
- 87ce505746f5: mptcp: never shrink offered window
- dac3ff7c87fe: mptcp: add more offered MIBs counter
- Results: 9709ecfa06e8..5b7d53ca0bd1 (export)
Builds and tests are now in progress:
https://cirrus-ci.com/github/multipath-tcp/mptcp_net-next/export/20220422T151603
https://github.com/multipath-tcp/mptcp_net-next/actions/workflows/build-validation.yml?query=branch:export
Cheers,
Matt
Signature · 2 lines
Tessares | Belgium | Hybrid Access Solutions www.tessares.net