Hi,
Perhaps I should be asking this in a regular newsgroup about samba,
but since it used to work, I will try it here first.
After a power-dip SAMBA does not work anymore.
If I look at the file /opt/var/samba/logs/log.nmbd I see:
(date) lib/util/ sock.c:open_socket_in(768)
open_socket_in(): setsockopt: SO_REUSEADDR = True on port 137 failed
with error = Operation not supported
(date) lib/util/ sock.c:open_socket_in(777))
open_socket_in(): setsockopt: SO_REUSEADDR = True on port 137 failed
with error = Operation not supported
(date) lib/util/ sock.c:open_socket_in(789))
bind failed on port 137 socket_addr = 10.2.8.92.
Error = Address already in use
And this about once every second.
What is wrong here ?
I know from other systems that if a service is aborted while a client
is connected, the address is in use for some time. But in this case
even rebooting does not help.
HELP !
(perhaps off-topic, but related: How do I perform a filesystem check
to fix corrupted files (Samba was complaining about a corrupted lock.
I removed the file and a new one was created)
Johan
The tiny stack doesn’t support SO_REUSEADDR on udp
sockets; however I thought samba used tcp by default.
You can check which stack you’re using with ‘pidin me’,
looking at the io-net process. npm-ttcpip.so is the
tiny stack, npm-tcpip.so is the full stack.
-seanb
Johan Siegers <johan.siegers@asm.com> wrote:
: Hi,
: Perhaps I should be asking this in a regular newsgroup about samba,
: but since it used to work, I will try it here first.
: After a power-dip SAMBA does not work anymore.
: If I look at the file /opt/var/samba/logs/log.nmbd I see:
: (date) lib/util/ sock.c:open_socket_in(768)
: open_socket_in(): setsockopt: SO_REUSEADDR = True on port 137 failed
: with error = Operation not supported
: (date) lib/util/ sock.c:open_socket_in(777))
: open_socket_in(): setsockopt: SO_REUSEADDR = True on port 137 failed
: with error = Operation not supported
: (date) lib/util/ sock.c:open_socket_in(789))
: bind failed on port 137 socket_addr = 10.2.8.92.
: Error = Address already in use
: And this about once every second.
: What is wrong here ?
: I know from other systems that if a service is aborted while a client
: is connected, the address is in use for some time. But in this case
: even rebooting does not help.
: HELP !
: (perhaps off-topic, but related: How do I perform a filesystem check
: to fix corrupted files (Samba was complaining about a corrupted lock.
look at the ‘chkfsys’ utility.
: I removed the file and a new one was created)
: Johan
Hi Sean,
Thanks. The problem was the tiny tcpip stack
I changed it so that I use the normal tcpip stack, and now nmbd does
work. (previously I could not connect to the computer by its name)
Johan
The tiny stack doesn’t support SO_REUSEADDR on udp
sockets; however I thought samba used tcp by default.
You can check which stack you’re using with ‘pidin me’,
looking at the io-net process. npm-ttcpip.so is the
tiny stack, npm-tcpip.so is the full stack.
-seanb
Johan Siegers <> johan.siegers@asm.com> > wrote:
: Hi,
: Perhaps I should be asking this in a regular newsgroup about samba,
: but since it used to work, I will try it here first.
: After a power-dip SAMBA does not work anymore.
: If I look at the file /opt/var/samba/logs/log.nmbd I see:
: (date) lib/util/ sock.c:open_socket_in(768)
: open_socket_in(): setsockopt: SO_REUSEADDR = True on port 137 failed
: with error = Operation not supported
: (date) lib/util/ sock.c:open_socket_in(777))
: open_socket_in(): setsockopt: SO_REUSEADDR = True on port 137 failed
: with error = Operation not supported
: (date) lib/util/ sock.c:open_socket_in(789))
: bind failed on port 137 socket_addr = 10.2.8.92.
: Error = Address already in use
: And this about once every second.
: What is wrong here ?
: I know from other systems that if a service is aborted while a client
: is connected, the address is in use for some time. But in this case
: even rebooting does not help.
: HELP !
: (perhaps off-topic, but related: How do I perform a filesystem check
: to fix corrupted files (Samba was complaining about a corrupted lock.
look at the ‘chkfsys’ utility.
: I removed the file and a new one was created)
: Johan