|
List Info
Thread: thinwire3 runaway process
|
|
| thinwire3 runaway process |
  United States |
2008-07-22 20:43:39 |
Hello,
I am not sure if anyone reads this list anymore, but if so,
here' a
question
When I launch k42 on mambo using k42console, thinwire gets
called as
follows:
thinwire3 localhost:44900 stdout 44902 44903 44904 44905 :
44909
44910 : 44917 44918 :
This creates several TCP sockets and binds them to the above
ports.
Sometimes I have to kill mambo ('systemsim-xxx'), and
restart it
again. When I restart it using 'k42console', I get a message
like this:
thinwire3: IO Channel 1 listening on port 44902 for stream
1.
thinwire3: IO Channel 2 listening on port 44903 for stream
2.
thinwire3: IO Channel 3 listening on port 44904 for stream
3.
thinwire3: IO Channel 4 listening on port 44905 for stream
4.
thinwire3: IO Channel 5 listening on port 44909 for stream
8.
thinwire3: IO Channel 6 listening on port 44910 for stream
9.
thinwire3: Retrying bind on port 44917.
thinwire3: Retrying bind on port 44917.
thinwire3: Retrying bind on port 44917.
...
thinwire3: Too many retries for bind on port 44917 for
stream 16
The problem is that the first mambo execution left a runaway
process:
leon phx1:~> sudo netstat -natp
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address Foreign
Address State PID/Program name
...
tcp 0 0 0.0.0.0:44917
0.0.0.0
LISTEN -
...
Unfortunately, there's no process id associated with this
TCP
connection, so I cannot kill it.
Any ideas on how I can kill this process (rebooting is not
an option),
or how to "overwrite" this connection with the new
one.
Also, it seems that SO_REUSEADDR is enabled in thinwire3.c:
SetSocketFlag(fd, SOL_SOCKET, SO_REUSEADDR);
Thanks!
Edgar.
--
Edgar Leon
Computer Science Dept.
University of New Mexico
http://www.cs.unm.edu/~le
on
_______________________________________________
K42-discussion mailing list
K42-discussion ozlabs.org
ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
|
|
| Re: thinwire3 runaway process |
  United States |
2008-07-25 17:21:46 |
Edgar A. Leon wrote:
> Hello,
>
> I am not sure if anyone reads this list anymore, but if
so, here' a
> question
>
> When I launch k42 on mambo using k42console, thinwire
gets called as
> follows:
>
> thinwire3 localhost:44900 stdout 44902 44903 44904
44905 : 44909 44910
> : 44917 44918 :
>
> This creates several TCP sockets and binds them to the
above ports.
>
> Sometimes I have to kill mambo ('systemsim-xxx'), and
restart it
> again. When I restart it using 'k42console', I get a
message like this:
>
> thinwire3: IO Channel 1 listening on port 44902 for
stream 1.
> thinwire3: IO Channel 2 listening on port 44903 for
stream 2.
> thinwire3: IO Channel 3 listening on port 44904 for
stream 3.
> thinwire3: IO Channel 4 listening on port 44905 for
stream 4.
> thinwire3: IO Channel 5 listening on port 44909 for
stream 8.
> thinwire3: IO Channel 6 listening on port 44910 for
stream 9.
> thinwire3: Retrying bind on port 44917.
> thinwire3: Retrying bind on port 44917.
> thinwire3: Retrying bind on port 44917.
> ...
> thinwire3: Too many retries for bind on port 44917 for
stream 16
>
> The problem is that the first mambo execution left a
runaway process:
>
> leon phx1:~> sudo netstat -natp
> Active Internet connections (servers and established)
> Proto Recv-Q Send-Q Local Address Foreign
> Address State PID/Program name
> ...
> tcp 0 0 0.0.0.0:44917
> 0.0.0.0
LISTEN -
> ...
>
> Unfortunately, there's no process id associated with
this TCP
> connection, so I cannot kill it.
>
> Any ideas on how I can kill this process (rebooting is
not an option),
> or how to "overwrite" this connection with
the new one.
>
> Also, it seems that SO_REUSEADDR is enabled in
thinwire3.c:
>
> SetSocketFlag(fd, SOL_SOCKET, SO_REUSEADDR);
>
>
> Thanks!
> Edgar.
>
> --
> Edgar Leon
> Computer Science Dept.
> University of New Mexico
> http://www.cs.unm.edu/~le
on
>
>
>
> _______________________________________________
> K42-discussion mailing list
> K42-discussion ozlabs.org
> ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
>
I am on the list but do not know much about programming. I
am hoping
others are watching and be able to help and with such help
enable to
learn a little more.
--
TomLeeM / BigWarpGuy * * * * * * * * * * * * * * * * * * * *
*
* * Director of Communications - * - Linkmaster * * * *
* * *
* * OS/2 Warp - eComStation Org * * OS/2 World * * * * * *
*
* * http://www.os2ecs.org * *
http://www.os2world.com
* *
* * Supporting Past OS/2 Users and Future eCS Users * * * *
* *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* *
* * * * http://home.comcast.
net/~tomleem * * * * * * * * * * *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* *
_______________________________________________
K42-discussion mailing list
K42-discussion ozlabs.org
ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
|
|
| Re: thinwire3 runaway process |
  United States |
2008-07-25 17:48:34 |
On Jul 22, 2008, at 9:43 PM, Edgar A. Leon wrote:
> Hello,
>
> I am not sure if anyone reads this list anymore, but if
so, here' a
> question
Hey Edgar,
I am but I'm not sure how much I can help you.
>
>
> When I launch k42 on mambo using k42console, thinwire
gets called as
> follows:
>
> thinwire3 localhost:44900 stdout 44902 44903 44904
44905 : 44909
> 44910 : 44917 44918 :
>
> This creates several TCP sockets and binds them to the
above ports.
>
> Sometimes I have to kill mambo ('systemsim-xxx'), and
restart it
> again. When I restart it using 'k42console', I get a
message like
> this:
>
> thinwire3: IO Channel 1 listening on port 44902 for
stream 1.
> thinwire3: IO Channel 2 listening on port 44903 for
stream 2.
> thinwire3: IO Channel 3 listening on port 44904 for
stream 3.
> thinwire3: IO Channel 4 listening on port 44905 for
stream 4.
> thinwire3: IO Channel 5 listening on port 44909 for
stream 8.
> thinwire3: IO Channel 6 listening on port 44910 for
stream 9.
> thinwire3: Retrying bind on port 44917.
> thinwire3: Retrying bind on port 44917.
> thinwire3: Retrying bind on port 44917.
> ...
> thinwire3: Too many retries for bind on port 44917 for
stream 16
>
> The problem is that the first mambo execution left a
runaway process:
>
> leon phx1:~> sudo netstat -natp
> Active Internet connections (servers and established)
> Proto Recv-Q Send-Q Local Address Foreign
> Address State PID/Program name
> ...
> tcp 0 0 0.0.0.0:44917
> 0.0.0.0
LISTEN -
> ...
>
This is very odd to have an un-named zombie, are you sure
about the
output?
How about lsof?
$ lsof -i:44917
or fuser?
$ fuser -n tcp 44917
The use fuser -k, that should kill it dead.
> Unfortunately, there's no process id associated with
this TCP
> connection, so I cannot kill it.
>
If you recall, we all had our own ksim scripts, mine
predates the
k42console abomination (hoping mxl is still watching) :-D
> Any ideas on how I can kill this process (rebooting is
not an
> option), or how to "overwrite" this
connection with the new one.
>
> Also, it seems that SO_REUSEADDR is enabled in
thinwire3.c:
>
> SetSocketFlag(fd, SOL_SOCKET, SO_REUSEADDR);
>
Good luck, let us know what worked.
-JX
>
> Thanks!
> Edgar.
>
> --
> Edgar Leon
> Computer Science Dept.
> University of New Mexico
> http://www.cs.unm.edu/~le
on
>
>
>
> _______________________________________________
> K42-discussion mailing list
> K42-discussion ozlabs.org
> ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
_______________________________________________
K42-discussion mailing list
K42-discussion ozlabs.org
ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
|
|
| Re: thinwire3 runaway process |
  United States |
2008-08-10 11:43:48 |
Hi Jimi,
> Hey Edgar,
> I am but I'm not sure how much I can help you.
Somehow I did not get this email until today (go figure...).
Knowing
that some people are still reading this list, I will send
some other
question I've have had a hard time fixing/understanding
(such as
debugging ...).
>> When I launch k42 on mambo using k42console,
thinwire gets called
>> as follows:
>>
>> thinwire3 localhost:44900 stdout 44902 44903 44904
44905 : 44909
>> 44910 : 44917 44918 :
>>
>> This creates several TCP sockets and binds them to
the above ports.
>>
>> Sometimes I have to kill mambo ('systemsim-xxx'),
and restart it
>> again. When I restart it using 'k42console', I get
a message like
>> this:
>>
>> thinwire3: IO Channel 1 listening on port 44902 for
stream 1.
>> thinwire3: IO Channel 2 listening on port 44903 for
stream 2.
>> thinwire3: IO Channel 3 listening on port 44904 for
stream 3.
>> thinwire3: IO Channel 4 listening on port 44905 for
stream 4.
>> thinwire3: IO Channel 5 listening on port 44909 for
stream 8.
>> thinwire3: IO Channel 6 listening on port 44910 for
stream 9.
>> thinwire3: Retrying bind on port 44917.
>> thinwire3: Retrying bind on port 44917.
>> thinwire3: Retrying bind on port 44917.
>> ...
>> thinwire3: Too many retries for bind on port 44917
for stream 16
>>
>> The problem is that the first mambo execution left
a runaway process:
>>
>> leon phx1:~> sudo netstat -natp
>> Active Internet connections (servers and
established)
>> Proto Recv-Q Send-Q Local Address
Foreign
>> Address State PID/Program name
>> ...
>> tcp 0 0 0.0.0.0:44917
>> 0.0.0.0
LISTEN -
>> ...
>>
>
> This is very odd to have an un-named zombie, are you
sure about the
> output?
> How about lsof?
> $ lsof -i:44917
> or fuser?
> $ fuser -n tcp 44917
> The use fuser -k, that should kill it dead.
I remember trying both lsof and fuser, and still did not get
a pid,
nor was I able to kill it.
Fortunately, I haven't seen this problem again. I ended up
rebooting
the machine and being careful about always checking that the
ports
were not in use before launching k42 console again.
Sometimes I have
to wait a few seconds to allow all the tcp connections to
die
completely.
>> Unfortunately, there's no process id associated
with this TCP
>> connection, so I cannot kill it.
>>
>
> If you recall, we all had our own ksim scripts, mine
predates the
> k42console abomination (hoping mxl is still watching)
:-D
>
>> Any ideas on how I can kill this process (rebooting
is not an
>> option), or how to "overwrite" this
connection with the new one.
>>
>> Also, it seems that SO_REUSEADDR is enabled in
thinwire3.c:
>>
>> SetSocketFlag(fd, SOL_SOCKET, SO_REUSEADDR);
>>
>
> Good luck, let us know what worked.
> -JX
Thanks!
Edgar.
--
Edgar Leon
Computer Science Dept.
University of New Mexico
http://www.cs.unm.edu/~le
on
_______________________________________________
K42-discussion mailing list
K42-discussion ozlabs.org
ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
|
|
| Re: thinwire3 runaway process |
  United States |
2008-08-10 14:02:39 |
Hey Edgar,
I have no idea for sure, but my first thought is TCP
time-wait delay. If
you want to track this down let me know I can give you some
ideas.
kurt
_______________________________________________
K42-discussion mailing list
K42-discussion ozlabs.org
ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
|
|
| Re: thinwire3 runaway process |
  United States |
2008-08-10 16:24:31 |
Hi Kurt,
Thanks. Yes, I saw the time-wait delay for a few seconds
right after I
killed thinwire and mambo. Before I launch k42console again,
I try to
wait until these connections are gone to avoid any
problems.
Cheers,
Edgar.
On Aug 10, 2008, at 1:02 PM, Kurt Ferreira wrote:
> Hey Edgar,
>
> I have no idea for sure, but my first thought is TCP
time-wait
> delay. If you want to track this down let me know I
can give you
> some ideas.
>
> kurt
> _______________________________________________
> K42-discussion mailing list
> K42-discussion ozlabs.org
> ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
--
Edgar Leon
Computer Science Dept.
University of New Mexico
http://www.cs.unm.edu/~le
on
_______________________________________________
K42-discussion mailing list
K42-discussion ozlabs.org
ht
tps://ozlabs.org/mailman/listinfo/k42-discussion
|
|
[1-6]
|
|