Prev: what happened to "text file busy" ?
Next: ld: fatal: relocations remain against allocatable but non-writable sections
From: Frank Cusack on 17 Nov 2006 11:42
On 17 Nov 2006 07:41:56 -0800 "AHinMaine" <aharrison(a)gmail.com> wrote:
> On Nov 17, 8:07 am, Michael Schreiber <nos...(a)schreiber.at> wrote:
>> hi,
>>
>> i have a problem with lsof under solaris 10 sparc:
>>
>> lsof -i |tail -2
>> httpd 29589 wwwrun 123u IPv4 0x300100cd600 0t0 TCP *:65535
>> (LISTEN)
>> httpd 29589 wwwrun 124u IPv4 0x3000fd8c500 0t0 TCP *:65535
>> (LISTEN)
>>
>> -> WHY IS THE PORT 80 NOT LISTET? *:65535 -> should be *:80 or *:http
>
> It is a vast assumption that the last two lines of your lsof output
> will be httpd using port 80. Instead, try lsof -i :80

Not at all. Look at the process name.

-frank
From: sil on 18 Nov 2006 01:53

Michael Schreiber wrote:
> hi,
>
> i have a problem with lsof under solaris 10 sparc:
>
> lsof -i |tail -2
> httpd 29589 wwwrun 123u IPv4 0x300100cd600 0t0 TCP *:65535
> (LISTEN)
> httpd 29589 wwwrun 124u IPv4 0x3000fd8c500 0t0 TCP *:65535
> (LISTEN)

Have you tried to see how Apache is configured by any chance

grep -i listen /path/to/apache/conf/httpd.conf

Have you tried telnetting to that machine on port 65535 if it is Apache
while in telnet you could type HEAD to make sure.

From: Dan Foster on 18 Nov 2006 02:23
In article <1163832807.684953.167920(a)m73g2000cwd.googlegroups.com>, sil <dsphunxion(a)gmail.com> wrote:
>
> Michael Schreiber wrote:
>> hi,
>>
>> i have a problem with lsof under solaris 10 sparc:
>>
>> lsof -i |tail -2
>> httpd 29589 wwwrun 123u IPv4 0x300100cd600 0t0 TCP *:65535
>> (LISTEN)
>> httpd 29589 wwwrun 124u IPv4 0x3000fd8c500 0t0 TCP *:65535
>> (LISTEN)
>
> Have you tried to see how Apache is configured by any chance
>
> grep -i listen /path/to/apache/conf/httpd.conf
>
> Have you tried telnetting to that machine on port 65535 if it is Apache
> while in telnet you could type HEAD to make sure.

The issue does not lie with Apache. Rather, it lies with lsof, as I can
reproduce it on my own fully patched Solaris 10/x86 Update 2 system:

# lsof | grep LISTEN
sendmail 386 root 4u IPv4 0xffffffff8187bd80 0t0 TCP *:65535 (LISTEN)
named 25636 dnsuser 23u IPv4 0xffffffff847561c0 0t0 TCP *:65535 (LISTEN)

Please observe that the above output claims that both sendmail and named
are listening on port 65535.

This is clearly impossible. Thus, it cannot be application
configuration, but rather, with lsof parsing kernel memory.

Casper Dik indicates that this would be fixed by recompiling lsof...
however, even when I do so with lsof 4.77 on Solaris 10/x86 Update 2 +
patches + SUNWgcc, still doesn't work. So... hmm. I'm not sure.

-Dan
From: abe on 18 Nov 2006 04:03

Dan Foster wrote:
> In article <1163832807.684953.167920(a)m73g2000cwd.googlegroups.com>, sil <dsphunxion(a)gmail.com> wrote:
> >
> > Michael Schreiber wrote:
> >> hi,
> >>
> >> i have a problem with lsof under solaris 10 sparc:
> >>
> >> lsof -i |tail -2
> >> httpd 29589 wwwrun 123u IPv4 0x300100cd600 0t0 TCP *:65535
> >> (LISTEN)
> >> httpd 29589 wwwrun 124u IPv4 0x3000fd8c500 0t0 TCP *:65535
> >> (LISTEN)
> >
> > Have you tried to see how Apache is configured by any chance
> >
> > grep -i listen /path/to/apache/conf/httpd.conf
> >
> > Have you tried telnetting to that machine on port 65535 if it is Apache
> > while in telnet you could type HEAD to make sure.
>
> The issue does not lie with Apache. Rather, it lies with lsof, as I can
> reproduce it on my own fully patched Solaris 10/x86 Update 2 system:
>
> # lsof | grep LISTEN
> sendmail 386 root 4u IPv4 0xffffffff8187bd80 0t0 TCP *:65535 (LISTEN)
> named 25636 dnsuser 23u IPv4 0xffffffff847561c0 0t0 TCP *:65535 (LISTEN)
>
> Please observe that the above output claims that both sendmail and named
> are listening on port 65535.
>
> This is clearly impossible. Thus, it cannot be application
> configuration, but rather, with lsof parsing kernel memory.
>
> Casper Dik indicates that this would be fixed by recompiling lsof...
> however, even when I do so with lsof 4.77 on Solaris 10/x86 Update 2 +
> patches + SUNWgcc, still doesn't work. So... hmm. I'm not sure.
>
> -Dan

Is always a good idea to look for lsof patches. There is a patch to
lsof 4.77 for this problem
It may be found on lsof.itap.purdue.edu in
pub/tools/unix/lsof/patches/4.77/solaris10-0606.patch.

Vic Abell, lsof author

From: Dan Foster on 18 Nov 2006 04:07
In article <1163840623.116755.148110(a)h48g2000cwc.googlegroups.com>, abe(a)purdue.edu <abe(a)purdue.edu> wrote:
>
> Is always a good idea to look for lsof patches.

Point well taken; mea culpa.

> There is a patch to lsof 4.77 for this problem

Ah, thank you very much -- most appreciated. I've always thought highly
of lsof for the decade I've been using it on multiple platforms.

-Dan
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: what happened to "text file busy" ?
Next: ld: fatal: relocations remain against allocatable but non-writable sections