On my computer, I noticed the following processes information:

$ pgrep -af f4f4a488c740adc8
3423257 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1487_f4f4a488c740adc80e1e4c43d81a14da.json -l 10014
3423263 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1487_f4f4a488c740adc80e1e4c43d81a14da.json -l 10014

As you can see, there are two processes listening on the same port. On the other hand, if I kill one of them, both will be terminated. Why do these processes work this way? Why are they listening on the same port at the same time? Any tips for this behavior will be appreciated.

Regards,
Zhao

On Sun, 26 Feb 2023 00:47:01 -0800 (PST), "hongy...@gmail.com"
<hongyi.zhao@gmail.com> wrote:

>On my computer, I noticed the following processes information:
>
> $ pgrep -af f4f4a488c740adc8
> 3423257 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1487_f4f4a488c740adc80e1e4c43d81a14da.json -l 10014
> 3423263 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1487_f4f4a488c740adc80e1e4c43d81a14da.json -l 10014
>
> As you can see, there are two processes listening
> on the same port. On the other hand, if I kill
> one of them, both will be terminated.

This pgrep output does not prove these sslocal processes are
listening on the same port. It just says they are started with
the same -l parameter.

Two processes _can_ listen on the same portnumber on _different
interfaces_.

> Why do these processes work this way?
> Why are they listening on the same port at the same time?

You'll have to study the documentation of sslocal, and/or ask
the author(s).

--
Kees Nuyt

On Sunday, February 26, 2023 at 11:14:21 PM UTC+8, Kees Nuyt wrote:
> On Sun, 26 Feb 2023 00:47:01 -0800 (PST), "hongy...@gmail.com"
> <hongy...@gmail.com> wrote:
>
> >On my computer, I noticed the following processes information:
> >
> > $ pgrep -af f4f4a488c740adc8
> > 3423257 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1487_f4f4a488c740adc80e1e4c43d81a14da.json -l 10014
> > 3423263 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1487_f4f4a488c740adc80e1e4c43d81a14da.json -l 10014
> >
> > As you can see, there are two processes listening
> > on the same port. On the other hand, if I kill
> > one of them, both will be terminated.
> This pgrep output does not prove these sslocal processes are
> listening on the same port. It just says they are started with
> the same -l parameter.
>
> Two processes _can_ listen on the same portnumber on _different
> interfaces_.

What might be the purpose of this? Can you give some similar examples?

> > Why do these processes work this way?
> > Why are they listening on the same port at the same time?
> You'll have to study the documentation of sslocal, and/or ask
> the author(s).
>
> --
> Kees Nuyt

Best,
Zhao

On Sun, 26 Feb 2023 07:46:20 -0800 (PST), "hongy...@gmail.com"
<hongyi.zhao@gmail.com> wrote:

> Kees Nuyt wrote:
>> Two processes _can_ listen on the same portnumber on _different
>> interfaces_.
>
> What might be the purpose of this? Can you give some similar examples?

I have no idea, but it is possible.
Check with
sudo netstat -lntp
what your sslocal processes are actually doing.
--
Kees Nuyt

On Monday, February 27, 2023 at 10:10:31 AM UTC+8, Kees Nuyt wrote:
> On Sun, 26 Feb 2023 07:46:20 -0800 (PST), "hongy...@gmail.com"
> <hongy...@gmail.com> wrote:
> > Kees Nuyt wrote:
> >> Two processes _can_ listen on the same portnumber on _different
> >> interfaces_.
> >
> > What might be the purpose of this? Can you give some similar examples?
> I have no idea, but it is possible.
> Check with
> sudo netstat -lntp
> what your sslocal processes are actually doing.

See the following:

werner@X10DAi:~$ sudo apt install net-tools
werner@X10DAi:~$ sudo netstat -lntp |grep 10014
tcp 0 0 0.0.0.0:10014 0.0.0.0:* LISTEN 239621/sslocal

In this way, only one process on each port returns in the command result.

> --
> Kees Nuyt

Regards,
Zhao

hongy...@gmail.com <hongyi.zhao@gmail.com> wrote:
>
>As you can see, there are two processes listening on the same port. On the other hand,
>if I kill one of them, both will be terminated. Why do these processes work this way?
>Why are they listening on the same port at the same time? Any tips for this behavior
>will be appreciated.
>

When you examine these two process ids and take into account each one's ppid
(parent process id), does it perhaps reveal these are a parent and child process?

--
::::::::::::: Greg Andrews ::::: gerg@panix.com :::::::::::::
I have a map of the United States that's actual size.
-- Steven Wright

On Thursday, March 2, 2023 at 2:37:58 PM UTC+8, gerg wrote:
> hongy...@gmail.com <hongy...@gmail.com> wrote:
> >
> >As you can see, there are two processes listening on the same port. On the other hand,
> >if I kill one of them, both will be terminated. Why do these processes work this way?
> >Why are they listening on the same port at the same time? Any tips for this behavior
> >will be appreciated.
> >
> When you examine these two process ids and take into account each one's ppid
> (parent process id), does it perhaps reveal these are a parent and child process?

They have the same pgid, sid, and sgid, as shown below:

werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
68311 67948 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
68318 68311 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014

> --
> ::::::::::::: Greg Andrews ::::: ge...@panix.com :::::::::::::
> I have a map of the United States that's actual size.
> -- Steven Wright

Zhao

Am 26.02.2023 schrieb "hongy...@gmail.com" <hongyi.zhao@gmail.com>:

> As you can see, there are two processes listening on the same port.

That is possible if they don't use the same IP.
So 2 different processes can listen on TCP [2001:db::123]:80 and TCP
[2001:db::4444]:80.

Use ss -ltun to find more information.

On Friday, March 3, 2023 at 4:51:22 PM UTC+8, Marco Moock wrote:
> Am 26.02.2023 schrieb "hongy...@gmail.com" <hongy...@gmail.com>:
>
> > As you can see, there are two processes listening on the same port.
> That is possible if they don't use the same IP.
> So 2 different processes can listen on TCP [2001:db::123]:80 and TCP
> [2001:db::4444]:80.
>
> Use ss -ltun to find more information.

See the following for more details:

werner@X10DAi:~$ ss -ltun | grep '10014'
udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:*
tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:*

Regards,
Zhao

On Thu, 2 Mar 2023 21:42:50 -0800 (PST), "hongy...@gmail.com"
<hongyi.zhao@gmail.com> wrote:

>
>They have the same pgid, sid, and sgid, as shown below:
>
>werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
> 68311 67948 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> 68318 68311 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014

So, process 68318 was started by 68311.

This is a common construct:
The parent process (68311 in this case) does not handle
anything, it just spawns a child process (68318). That child
does the heavy lifting. If the child crashes, the parent spawns
a new one. In that case the child listens on the port.

There are other constructs, where the parent listens on the port
and spawns one or more child ("worker") processes to handle
incomin requests. The parent just mamages the request queue.

I don't know what happens here. As I said before, you have to
consult the sslocal documentation.
Ok, enough keywords there for you to search for mor elaborate
descriptions.
--
Kees Nuyt

On Sat, 4 Mar 2023 03:51:40 -0800 (PST), "hongy...@gmail.com"
<hongyi.zhao@gmail.com> wrote:

>> Use ss -ltun to find more information.
>
> See the following for more details:
>
> werner@X10DAi:~$ ss -ltun | grep '10014'
> udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:*
> tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:*

Now, add the -p option to see if udp and tcp are served by the
same process:
sudo ss -ltunp
--
Kees Nuyt

On Saturday, March 4, 2023 at 8:29:07 PM UTC+8, Kees Nuyt wrote:
> On Thu, 2 Mar 2023 21:42:50 -0800 (PST), "hongy...@gmail.com"
> <hongy...@gmail.com> wrote:
>
> >
> >They have the same pgid, sid, and sgid, as shown below:
> >
> >werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
> > 68311 67948 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> > 68318 68311 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> So, process 68318 was started by 68311.
>
> This is a common construct:
> The parent process (68311 in this case) does not handle
> anything, it just spawns a child process (68318). That child
> does the heavy lifting. If the child crashes, the parent spawns
> a new one. In that case the child listens on the port.
>
> There are other constructs, where the parent listens on the port
> and spawns one or more child ("worker") processes to handle
> incomin requests. The parent just mamages the request queue.
>
> I don't know what happens here. As I said before, you have to
> consult the sslocal documentation.
> Ok, enough keywords there for you to search for mor elaborate
> descriptions.

Nice comments and explanations. Thank you very much.

> --
> Kees Nuyt

Zhao

On Saturday, March 4, 2023 at 8:36:09 PM UTC+8, Kees Nuyt wrote:
> On Sat, 4 Mar 2023 03:51:40 -0800 (PST), "hongy...@gmail.com"
> <hongy...@gmail.com> wrote:
>
> >> Use ss -ltun to find more information.
> >
> > See the following for more details:
> >
> > werner@X10DAi:~$ ss -ltun | grep '10014'
> > udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:*
> > tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:*
> Now, add the -p option to see if udp and tcp are served by the
> same process:
> sudo ss -ltunp

See the following:

werner@X10DAi:~$ ss -ltunp | grep '10014'
udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=4))
tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=3))
werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
3123709 3123347 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
3123715 3123709 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014

> --
> Kees Nuyt

Best,
Zhao

On Saturday, March 4, 2023 at 8:29:07 PM UTC+8, Kees Nuyt wrote:
> On Thu, 2 Mar 2023 21:42:50 -0800 (PST), "hongy...@gmail.com"
> <hongy...@gmail.com> wrote:
>
> >
> >They have the same pgid, sid, and sgid, as shown below:
> >
> >werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
> > 68311 67948 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> > 68318 68311 67948 64750 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> So, process 68318 was started by 68311.
>
> This is a common construct:
> The parent process (68311 in this case) does not handle
> anything, it just spawns a child process (68318). That child
> does the heavy lifting. If the child crashes, the parent spawns
> a new one. In that case the child listens on the port.

I noticed that if I killed the child process (68318 in this case), both processes will be terminated. I'm not sure whether this is consistent with your above analysis.
> There are other constructs, where the parent listens on the port
> and spawns one or more child ("worker") processes to handle
> incomin requests. The parent just mamages the request queue.
>
> I don't know what happens here. As I said before, you have to
> consult the sslocal documentation.
> Ok, enough keywords there for you to search for mor elaborate
> descriptions.
> --
> Kees Nuyt

On Sat, 4 Mar 2023 04:45:24 -0800 (PST), "hongy...@gmail.com"
<hongyi.zhao@gmail.com> wrote:

>> Now, add the -p option to see if udp and tcp are served by the
>> same process:
>> sudo ss -ltunp
>
>See the following:
>
> werner@X10DAi:~$ ss -ltunp | grep '10014'
> udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=4))
> tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=3))
> werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
> 3123709 3123347 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> 3123715 3123709 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014

Ok, there you are, that's clear. And what does the sslocal docs
tell you?
--
Kees Nuyt

Port numbers are not the same thing as sockets.

You can easily have different processes listen to the same port number
on different protocols (UDP vs TCP) and / or different IPs.

On 3/4/23 4:51 AM, hongy...@gmail.com wrote:
> werner@X10DAi:~$ ss -ltun | grep '10014'
> udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:*
> tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:*

This shows two different sockets. Both are listening on port 10014 on
all IPs. However one is on UDP and the other is on TCP.

It is also possible for a process to listen to multiple sockets.

There are also some provisions to pass a connection on a socket from a
parent process to a child process.

I think that there are some VERY SPECIAL things that can be done to have
two separate processes listen to the same socket (protocol / address /
port) but it requires VERY SPECIAL provisions including both processes
have to explicitly support it and I believe that the kernel has to
support it. -- I can't think of an example where I've seen two
different and unrelated processes listening on the same socket.

N.B. that things like `sslh` don't violate this precept because sslh is
the process listening on the socket and passing / proxying traffic to
the other processes listening on different sockets.

--
Grant. . . .
unix || die

On Monday, March 6, 2023 at 2:01:36 AM UTC+8, Kees Nuyt wrote:
> On Sat, 4 Mar 2023 04:45:24 -0800 (PST), "hongy...@gmail.com"
> <hongy...@gmail.com> wrote:
>
> >> Now, add the -p option to see if udp and tcp are served by the
> >> same process:
> >> sudo ss -ltunp
> >
> >See the following:
> >
> > werner@X10DAi:~$ ss -ltunp | grep '10014'
> > udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=4))
> > tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=3))
> > werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
> > 3123709 3123347 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> > 3123715 3123709 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> Ok, there you are, that's clear. And what does the sslocal docs
> tell you?

TBF, I don't know. This is a program that stopped developing many years ago, and its documentation is not friendly.

> --
> Kees Nuyt

Zhao

On Mon, 6 Mar 2023 00:05:38 -0800 (PST), "hongy...@gmail.com"
<hongyi.zhao@gmail.com> wrote:

>On Monday, March 6, 2023 at 2:01:36?AM UTC+8, Kees Nuyt wrote:
>> On Sat, 4 Mar 2023 04:45:24 -0800 (PST), "hongy...@gmail.com"
>> <hongy...@gmail.com> wrote:
>>
>> >> Now, add the -p option to see if udp and tcp are served by the
>> >> same process:
>> >> sudo ss -ltunp
>> >
>> >See the following:
>> >
>> > werner@X10DAi:~$ ss -ltunp | grep '10014'
>> > udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=4))
>> > tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=3))
>> > werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
>> > 3123709 3123347 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
>> > 3123715 3123709 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
>> Ok, there you are, that's clear. And what does the sslocal docs
>> tell you?
>
> TBF, I don't know. This is a program that stopped developing
> many years ago, and its documentation is not friendly.

It was brought to my attention that this subject is off-topic in
comp.unix.shell , so I declare this END-OF-THREAD
--
Kees Nuyt

On Monday, March 6, 2023 at 10:14:46 AM UTC-6, Kees Nuyt wrote:
> On Mon, 6 Mar 2023 00:05:38 -0800 (PST), "hongy...@gmail.com"
> <hongy...@gmail.com> wrote:
> >On Monday, March 6, 2023 at 2:01:36?AM UTC+8, Kees Nuyt wrote:
> >> On Sat, 4 Mar 2023 04:45:24 -0800 (PST), "hongy...@gmail.com"
> >> <hongy...@gmail.com> wrote:
> >>
> >> >> Now, add the -p option to see if udp and tcp are served by the
> >> >> same process:
> >> >> sudo ss -ltunp
> >> >
> >> >See the following:
> >> >
> >> > werner@X10DAi:~$ ss -ltunp | grep '10014'
> >> > udp UNCONN 0 0 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=4))
> >> > tcp LISTEN 0 1024 0.0.0.0:10014 0.0.0.0:* users:(("sslocal",pid=3123715,fd=3))
> >> > werner@X10DAi:~$ ps xao pid,ppid,pgid,sid,sgid,command | grep ' -l 10014' |grep -v grep
> >> > 3123709 3123347 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> >> > 3123715 3123709 3123347 3119533 1000 sslocal -c /home/werner/Public/anti-gfw/conf/ssr2json/ssr/US_ssr_eebzkdao.nodelist.xyz_1381_1421cc147ea4bb31049aea01851aac7e.json -l 10014
> >> Ok, there you are, that's clear. And what does the sslocal docs
> >> tell you?
> >
> > TBF, I don't know. This is a program that stopped developing
> > many years ago, and its documentation is not friendly.
> It was brought to my attention that this subject is off-topic in
> comp.unix.shell , so I declare this END-OF-THREAD
> --
> Kees Nuyt
fggggggtguttututttut8tttftytttffyy