Re: How long TCP state change from SYN_RECV to ESTABLISHED should take?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



When I tested on INITIATOR side, transition between SYN_SENT -> ESTABLISHED was near RTT, but on RECEIVER side it was very low, as i showed before, but it should be the same as on INITIATOR side. Is that right?
I just want to understand where to look at, if it's not netfilter.. Maybe eBPF or Linux networking?



09.06.2019, 22:48, "zrm" <zrm@xxxxxxxxxxxxxxx>:
> On 6/9/19 15:04, iam@xxxxxxxxxxx wrote:
>>  While playing with eBPF i tried tcpstates.py (https://github.com/iovisor/bcc/blob/master/tools/tcpstates.py) and noticed very low time for TCP state transition SYN_RECV -> ESTABLISHED
>>
>>  My numbers from tests:
>>
>>  SYN_RECV -> ESTABLISHED 0.015
>>  SYN_RECV -> ESTABLISHED 0.017
>>  SYN_RECV -> ESTABLISHED 0.051
>>
>>   From handshake diagram(https://user-images.githubusercontent.com/1006307/58944706-0ffdb580-878b-11e9-95d3-8e7a4f85d8b0.png) it looks like that transition time from SYN_RECV to ESTABLISHED should be near RTT between hosts?
>>
>>  TCP Fast Open turned off on host.
>>
>>  OS: Ubuntu 18.10 (GNU/Linux 4.18.0-21-generic x86_64)
>>
>>  tcpdump -tttttv
>>
>>    00:00:00.000000 IP (tos 0x28, ttl 49, id 280, offset 0, flags [DF], proto TCP (6), length 60)
>>       dst_host > src_host: Flags [S], cksum 0x46ae (correct), seq 4063608731, win 29200, options [mss 1460,sackOK,TS val 332512899 ecr 0,nop,wscale 6], length 0
>>    00:00:00.000071 IP (tos 0x0, ttl 64, id 0, offset 0, flags [DF], proto TCP (6), length 60)
>>       src_host > dst_host: Flags [S.], cksum 0x84c2 (incorrect -> 0x3c01), seq 1516815880, ack 4063608732, win 28960, options [mss 1460,sackOK,TS val 2341429130 ecr 332512899,nop,wscale 7], length 0
>>    00:00:00.079183 IP (tos 0x28, ttl 49, id 281, offset 0, flags [DF], proto TCP (6), length 52)
>>       dst_host > src_host: Flags [.], cksum 0xda11 (correct), ack 1, win 457, options [nop,nop,TS val 332512918 ecr 2341429130], length 0
>>    00:00:00.093311 IP (tos 0x0, ttl 64, id 60333, offset 0, flags [DF], proto TCP (6), length 93)
>>       src_host > dst_host: Flags [P.], cksum 0x84e3 (incorrect -> 0xb341), seq 1:42, ack 1, win 227, options [nop,nop,TS val 2341429223 ecr 332512918], length 41
>>  such low numbers are between syn_recv and syn+ack send, but between syn_recv and established it should be 00:00:00.079183
>>
>>  ping between hosts in this tcpdump is around 83ms
>>
>>  Looks like nf_conntrack is doing this:
>>
>>  https://github.com/torvalds/linux/blob/80f232121b69cc69a31ccb2b38c1665d770b0710/net/netfilter/nf_conntrack_proto_tcp.c#L1033-L1037
>>
>>  Is it bug or i'm wrong?
>
> See how on your TCP diagram there is ESTABLISHED on both sides?
> Conntrack isn't the initiator or the receiver, it's on the path between
> them (even if it's physically on the same machine as the receiver), and
> it transitions to ESTABLISHED as soon as it sees SYN-ACK from the
> receiver, the same as the initiator does.



[Index of Archives]     [Linux Netfilter Development]     [Linux Kernel Networking Development]     [Netem]     [Berkeley Packet Filter]     [Linux Kernel Development]     [Advanced Routing & Traffice Control]     [Bugtraq]

  Powered by Linux