From: Ian Kent <raven@themaw.net>
To: Julian Hagenauer <chaosbringer@gmx.de>
Cc: autofs@linux.kernel.org
Subject: Re: NFS+AUTOFS: Strange connections every 2 minutes
Date: Tue, 07 Nov 2006 20:06:46 +0800 [thread overview]
Message-ID: <1162901206.18975.40.camel@localhost> (raw)
In-Reply-To: <20061107105134.3618b93b@notebook.chaosbringer.de>
On Tue, 2006-11-07 at 10:51 +0100, Julian Hagenauer wrote:
> Hi,
> i think it is a autofs and not nfs-related problem, because it does not occur,if i have stopped the automounter.
> I set up a nfs-share to get mounted with tcp by the automounter, but even if i do not access the share on client-side, tcp-connections still persist.
It's way to early to start looking at packets.
In any case autofs uses mount(8) and so a packet trace is usually not
useful.
> Those are the connections that permanently persist:
What do you mean permanently persist.
They are connections that have been closed and are going away after
TIME_WAIT seconds.
> tcp 6 13 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=933 dport=111 packets=6 bytes=364 src=192.168.1.9 dst=192.168.1.4 sport=111 dport=933 packets=4 bytes=408 [ASSURED] use=1
> tcp 6 16 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=1023 dport=2049 packets=10 bytes=896 src=192.168.1.9 dst=192.168.1.4 sport=2049 dport=1023 packets=7 bytes=584 [ASSURED] use=1
> tcp 6 13 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=934 dport=878 packets=6 bytes=412 src=192.168.1.9 dst=192.168.1.4 sport=878 dport=934 packets=4 bytes=280 [ASSURED] use=1
> tcp 6 16 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=937 dport=111 packets=6 bytes=380 src=192.168.1.9 dst=192.168.1.4 sport=111 dport=937 packets=4 bytes=248 [ASSURED] use=1
> tcp 6 13 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=936 dport=111 packets=6 bytes=380 src=192.168.1.9 dst=192.168.1.4 sport=111 dport=936 packets=4 bytes=248 [ASSURED] use=1
> tcp 6 16 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=938 dport=878 packets=6 bytes=412 src=192.168.1.9 dst=192.168.1.4 sport=878 dport=938 packets=4 bytes=244 [ASSURED] use=1
>
> If the TIME_WAIT period is over, new TCP-Connections are done and i have again 6 connections in Time_Wait state:
If you have a process that is periodically trying to access something in
an automount managed mount then this is exactly what you will see.
>
> tcp 6 43 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=1023 dport=2049 packets=10 bytes=896 src=192.168.1.9 dst=192.168.1.4 sport=2049 dport=1023 packets=7 bytes=584 [ASSURED] use=1
> tcp 6 40 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=953 dport=111 packets=6 bytes=380 src=192.168.1.9 dst=192.168.1.4 sport=111 dport=953 packets=4 bytes=248 [ASSURED] use=1
> tcp 6 43 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=955 dport=878 packets=6 bytes=412 src=192.168.1.9 dst=192.168.1.4 sport=878 dport=955 packets=4 bytes=244 [ASSURED] use=1
> tcp 6 40 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=951 dport=878 packets=6 bytes=412 src=192.168.1.9 dst=192.168.1.4 sport=878 dport=951 packets=4 bytes=280 [ASSURED] use=1
> tcp 6 43 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=954 dport=111 packets=6 bytes=380 src=192.168.1.9 dst=192.168.1.4 sport=111 dport=954 packets=4 bytes=248 [ASSURED] use=1
> tcp 6 40 TIME_WAIT src=192.168.1.4 dst=192.168.1.9 sport=950 dport=111 packets=6 bytes=364 src=192.168.1.9 dst=192.168.1.4 sport=111 dport=950 packets=4 bytes=408 [ASSUERD] use=1
>
snip ...
> Again, if i stop the automounter the TIME-WAIT-connections are closed and no new ones are made.
> I want no connection to persist, if the automounter has not mounted anything.
> How can i accomblish that?
It could be something else but experience so far has shown that this is
caused by some process continually accessing mount points. Unless there
is other evidence I can only recommend you find the process that's
scanning the directory tree or trying to access your automount
directories and fix it.
Ian
next prev parent reply other threads:[~2006-11-07 12:06 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-11-07 9:51 NFS+AUTOFS: Strange connections every 2 minutes Julian Hagenauer
2006-11-07 12:06 ` Ian Kent [this message]
2006-11-07 12:36 ` Julian Hagenauer
2006-11-07 12:46 ` Ian Kent
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1162901206.18975.40.camel@localhost \
--to=raven@themaw.net \
--cc=autofs@linux.kernel.org \
--cc=chaosbringer@gmx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.