echo 0 > /proc/sys/net/ipv4/tcp_timestamps per [session, reboot, connection]?

echo 0 > /proc/sys/net/ipv4/tcp_timestamps per [session, reboot, connection]?

Post by Dan Jacobso » Sat, 06 Oct 2001 07:25:19



excuse me but does
echo 0 > /proc/sys/net/ipv4/tcp_timestamps
need to be done after each [session, reboot, connection]? and where
should I put it [/etc/ppp/ip-up?] to have effect?

I'm trying to solve
http://www.tuxedo.org/~esr/fetchmail/fetchmail-FAQ.html#R9
R9. Fetchmail is timing out after fetching certain messages but before
deleting them
causing:
O9. Why does fetchmail keep retrieving the same messages over and
over?
--
http://www.geocities.com/jidanni Tel+886-4-25854780

 
 
 

echo 0 > /proc/sys/net/ipv4/tcp_timestamps per [session, reboot, connection]?

Post by Andrei Ivan » Sat, 06 Oct 2001 07:59:06



Quote:> excuse me but does
> echo 0 > /proc/sys/net/ipv4/tcp_timestamps
> need to be done after each [session, reboot, connection]? and where
> should I put it [/etc/ppp/ip-up?] to have effect?

It depends on distribution you're using. With RedHat 6.2 and newer
you should put "net.ipv4.tcp_timestamps = 0" line into
/etc/sysctl.conf file. With other distros you might put the "echo ..."
command into /etc/rc.d/rc.local, or its equivalent (this will work
for all RedHat versions too).

--
andrei

 
 
 

echo 0 > /proc/sys/net/ipv4/tcp_timestamps per [session, reboot, connection]?

Post by Karl Heye » Sat, 06 Oct 2001 09:09:55




> excuse me but does
> echo 0 > /proc/sys/net/ipv4/tcp_timestamps need to be done after each
> [session, reboot, connection]? and where should I put it [/etc/ppp/ip-up?]
> to have effect?

ip-up gets called when the ppp has been established.  Files under /proc
represent the current status of the kernel. It defaults to 1 but I believe
there has been some cases of routers and OS (I think windows) etc not
handling these options properly possibly giving the effect your seeing.

Quote:

> I'm trying to solve
> http://www.tuxedo.org/~esr/fetchmail/fetchmail-FAQ.html#R9 R9. Fetchmail is
> timing out after fetching certain messages but before deleting them causing:
> O9. Why does fetchmail keep retrieving the same messages over and over?

It it solves the problem your experiencing then fine.

karl.

 
 
 

echo 0 > /proc/sys/net/ipv4/tcp_timestamps per [session, reboot, connection]?

Post by Dan Jacobso » Fri, 12 Oct 2001 08:48:50


fetchmail[1245]: 5.5.2 querying pop.mail.yahoo.com.tw (protocol POP3)
fetchmail[1245]: POP3< +OK hello from popgate
fetchmail[1245]: POP3> USER jidannikimo^M
fetchmail[1245]: POP3< +OK password required.
fetchmail[1245]: POP3> PASS *^M
fetchmail[1245]: POP3< +OK maildrop ready, 0 messages (0 octets) (10058

[Hmmm, it already says there is no messages... maybe we could "quit
while still ahead"...]

fetchmail[1245]: selecting or re-polling default folder
fetchmail[1245]: POP3> STAT
fetchmail[1245]: socket error while fetching from pop.mail.yahoo.com.tw

Blast.  Even though I've done
echo 0 > /proc/sys/net/ipv4/tcp_timestamps
in the rc.local file, I'm still getting this :-(

fetchmail[1245]: Query status=2 (SOCKET)
fetchmail[1245]: Writing fetchids file.

[hmm, why write when didn't fetch anything?]
Anyway, I'm trying to solve
http://www.tuxedo.org/~esr/fetchmail/fetchmail-FAQ.html#O9
O9. Why does fetchmail keep retrieving the same messages over and over?

I've now even set maxfetch to 1...
--
http://www.geocities.com/jidanni/ Tel+886-4-25854780 ?n|