IP bandwidth consumption logging
Hello list.I'm looking for a little tool that will provide me with, say, daily summaries of how much traffic was on each TCP and UDP port. A bit like the rather nifty iptraf but not real-time.I want this because my server is capped to fairly small amounts of bandwidth each month and a friend has asked me to install teamspeak, a VoIP app. I want to be sure teamspeak isn't going to drive my bandwidth through the roof! Does anyone know of any software like this? Seems like the sort of thing that should be easily apt-gettable but I can't find anything either in there or google.Cheers,Richard
Re: ISDN diald: router setup tale of woe
Sorry about delay in replying to this, Tony, and thanks for your help so
far.
On Sat, Feb 16, 2002 at 08:06:09PM +0100, Tony Crawford wrote:
Richard Gaywood wrote (on 16 Feb 2002 at 18:37):
On Sat, Feb 16, 2002 at 07:21:34PM +0100, Tony Crawford wrote:
Richard Gaywood wrote (on 16 Feb 2002 at 18:05):
Have you read in the man page about the demand and persist
options?
man ipppd reports persist as obsolete in ipppd; it'd get me into
trouble with my ISP anyway, although I can see what you're saying.
It doesn't mention a demand option at all.
Yes, sorry, demand is a pppd option. What you want for ipppd is
dialmode=auto in /etc/isdn/device/ippp0. (Do you have lots of
nice explanatory comments in that file?)
Rggght. Because of this weird-assed config script I've used, most of
the config is tucked away in /etc/isdn/accounts/conf-MYISP. However, I
can boot the router, run /etc/init.d/isdn, and then do this:
[/etc/isdn]# isdnctrl writeconf foo
Will overwrite section `ISDNCTRL'!
ISDN Configuration written to foo.
[/etc/isdn]# cat foo
###
#
# File foo generated by isdnctrl 3.0beta2 on Thu Feb 21 22:47:35 2002
#
###
[ISDNCTRL]
INTERFACES = {
[INTERFACE]
NAME = ippp0
EAZ = 01495244285
PHONE_OUT = 08089916001 08089916001
SECURE = on
DIALMODE = auto
DIALMAX = 10
HUPTIMEOUT = 7200
IHUP = off
CHARGEHUP = off
L2_PROT = hdlc
L3_PROT = trans
ENCAP = syncppp
PPPBIND = 0
}
So it looks like it's in auto dialmode already. Another piece of the
puzzle that might be relevent:
[/etc/isdn]# route
Kernel IP routing table
Destination Gateway Genmask Flags Metric RefUse Iface
192.168.0.0 * 255.255.255.0 U 0 00 eth0
195.92.66.0 * 255.255.255.0 U 0 00 ippp0
default cougar.lns.watf 0.0.0.0 UG0 00 ippp0
[/etc/isdn]# isdnctrl hangup ippp0
ippp0 hung up
[/etc/isdn]# route
Kernel IP routing table
Destination Gateway Genmask Flags Metric RefUse Iface
192.168.0.0 * 255.255.255.0 U 0 00 eth0
195.92.66.0 * 255.255.255.0 U 0 00 ippp0
[/etc/isdn]# isdnctrl dial ippp0
Dialing of ippp0 triggered
[/etc/isdn]# route
Kernel IP routing table
Destination Gateway Genmask Flags Metric RefUse Iface
192.168.0.0 * 255.255.255.0 U 0 00 eth0
195.92.66.0 * 255.255.255.0 U 0 00 ippp0
default * 0.0.0.0 U 0 00 ippp0
Am I right in theorising the route is getting messed up?
That should make the ISDN line dial whenever packets are sent to the
interface ippp0. In order for packets to get sent there, you usually
want to set a default route--your handy-dandy config scripts may have
done this for you; verify using route -n.
Aha. Thought so. I'll have another comb through the config and try and
resolve where the default route gets deleted. What should happen to it
when ippp0 hangs up or the remote caller hangs up? I'm guessing it
should stay in place?
Also, the isdnutils package sets up scripts in /etc/ppp/ip-up.d/ and
...ip-down.d. You can amend these scripts to activate firewall rules
whenever you ISDN line is up, or to correct the routing. (There was a
time when the default route had to be restored in ip- down after every
hangup.) These scripts are also full of helpful comments.
A, right. Thanks for your help; hopefully I'll be able to get this
licked now. I'll let you know :oD
--
Got you in my sites and I'm gonna make you bleed, | -=R=-
Get used to hearing this, `you have lost the lead.' |web: http://
-- MC Hawking |fscked.co.uk
|icq:66545079
ISDN diald: router setup tale of woe
Hi. I'm having massive grief configuring a standalone ISDN router and I could desperately use some pointers. Quick recap of the story so far: I started with a CD-based install of 2.2r5, then cast around looking for ways to configure the ISDN. I've built routers before but never around dialup and I knew very little about ISDN, so the default Debian configure tool confused me. Google turned up a neat script at www.manna.nl which asked me questions in a lovely hand-holding manner and set everything up. Once this script was done, I had a init.d service called isdn that, when started, does *everything* to configure the line. I have this service to start at my default runlevel and on the router's console, I can type 'isdnctrl dial ippp0' immediately after booting and my net connection is instantly up. I figured this meant I'd broken the back of the problem. I have all the rest of the stuff I need-- masquerading, DHCP server for local LAN, BIND, etc-- up and running. The last piece of the puzzle is making the router dial out on demand but I've just spent nearly a full day banging my head at the diald docs trying to make it co-operate. I only need it to run isdnctrl dial and isdnctrl hangup to bring the link up and down, but it's just not working. Starting diald causes /var/log/syslog to fill with messages about modules tap0-15 not existing; the diald docs mention that I should have ethertap compiled in the kernel but modconf doesn't seem to know about it. Even though diald shows up on ps aux, it doesn't seem to ever do anything (pinging out etc doesn't make it dial) and route just shows my LAN route. I'm presuming either route or ifconfig should show up the SLIP loopback link diald uses. Meanwhile, I'm looking at various diald docs and walkthroughs and getting ever more confused because most of them assume normal dialup and no two of them seem to do things the same way. It's possible I'm supposed to be editing things under /etc/ppp but then again, maybe not because I'm using ipppd; I'm so confused now, I really can't tell any more! Any insight into how I can get this working is very much appreciated. I'd like to think I know what I'm doing with Debian after a year of use but this really has me frustrated. Here's the diald config I've put^Whacked^Wcludged together so far: = diald.conf === #fifo /var/run/diald/diald.fifo mode dev connect sh /etc/diald/connect device ippp0 lock crtscts # network options local 192.168.0.1 remote 195.92.66.181 #dynamic #defaultroute #pppd-options asyncmap 0 #include /etc/diald/standard.filter ip-up /etc/diald/ip-up addroute /etc/diald/addroute ip-down /etc/diald/ip-down delroute /etc/diald/delroute = connect == /etc/init.d/isdn restart = ip-up /usr/sbin/isdnctrl dial ippp0 = ip-down == /usr/sbin/isdnctrl hangup ippp0 = addroute = #!/bin/sh logger -p local1.info $0 $* #. /etc/diald/config # Setup the Link #/sbin/isdnctrl addphone $1 out $PHONE #/sbin/isdnctrl dial $1 # Setup Routes # /sbin/route add janus $1 /sbin/route add default $1 /sbin/route add $4 $1 exit 0 = delroute = #!/bin/sh . /etc/diald/config logger -p local1.info $0 $* # Delete The routes from the Link /sbin/route del default /sbin/route del $4 # Hangup the Linke #/sbin/isdnctrl delphone $1 out $PHONE #/sbin/isdnctrl delphone $1 out $PHONE #/sbin/isdnctrl delphone $1 out $PHONE /sbin/isdnctrl hangup $1 exit 0 diald.options is empty. Thanks in advance for anyone who actually read all that. -- He couldn't get a clue during the clue mating season in a field| -=R=- full of horny clues if he smeared his body with clue musk and did |web: http:// the clue mating dance. |fscked.co.uk |icq:66545079
Re: ISDN diald: router setup tale of woe
On Sat, Feb 16, 2002 at 07:21:34PM +0100, Tony Crawford wrote: Richard Gaywood wrote (on 16 Feb 2002 at 18:05): Hi. I'm having massive grief configuring a standalone ISDN router and I could desperately use some pointers. Are you using pppd or ipppd? ipppd. Is this a bad idea? What version? Default potato -- i2.2 patchlevel 10anubis. Although man ipppd reports 2.2.9, oddly. Have you read in the man page about the demand and persist options? man ipppd reports persist as obsolete in ipppd; it'd get me into trouble with my ISP anyway, although I can see what you're saying. It doesn't mention a demand option at all. -- I actually hate every piece of hardware invented apart from the | -=R=- ones I own, because I'm infallible and could never have misjudged |web: http:// any purchases so all my hardware is gr8 and all yours is fuvg.|fscked.co.uk -- doki in [EMAIL PROTECTED][ROT13]|icq:66545079
Debian keeps hanging!
My Debian testing install keeps locking up solid under X and it's driving me crazy. At apparantly random intervals - sometimes hours, sometimes minutes - the entire machine just comes to a screeching halt. Not even Magic SysRq can persuade it to do anything afterwards. I'm getting real tired of fsck bootups... I used to have a problem with the Xserver dying at random on my Xinerama setup, but pulling the second graphics card (an S3) fixed that. Now I just get this hanging. Pertinent statistics: - recently 'apt-get dist-upgrade'd testing distribution, with the normal X 4.0.3 - custom compiled 2.4.7 kernel (system has been unstable with 2.4.1 onwards, although I think it might be getting worse) - Gnome 1.4 (tried both Ximian and Debian builds) - also seen instability under KDE - nothing at all useful goes to /var/log/syslog before the hang - Previous installs of Redhat 7 and Win2k on the same hardware don't seem to be affected - more system specs are available at http://www.fscked.co.uk/stuff/hardware.html What I've tried: - backing off my previously overclocked CPU to default speed and voltage - checked cooling (four system fans, I think it's OK) - changed from nVidia drivers to the default Xfree nv driver for my Geforce 256 - running the SuSE memtest-86 ram tester on my memory (it checked OK) Please, can anyone with any thoughts at all, no matter how tenuous, help me out here. I don't even know where to look for more problems right now. -- -=R=- print join ,reverse split /\s+/,'hacker. Perl another Just',\n; Web: http://www.fscked.co.uk ICQ: 66545073
Massive problems with Gnome/KDE/kdm/gdm Xinerama
Heh, I think my subject line has covered all the bases :o) Up until recently, I was happy using gdm and Gnome, then having managed to free up a second monitor I thought I'd give Xinerama a whirl. Everything was OK... sort of; it worked, but X suddently acquired an unpleasant habit of locking solid at random, to the point where even Magic SysReq doesn't seem to work any more. At other times, it would just suddenly die and kick me back out to gdm, with a log file message saying: Jul 22 22:15:28 glyndwr gnome-name-server[877]: input condition is: 0x11, exiting (that's the last entry in syslog) That brings me onto the next problem... upon X server death (via ctrl-alt-backspace, logout, or this random crash) gdm would fail to come up about 4/5 of the time, resulting in a hard lock and yet another fsck bootup. This is getting rather tedious, and I've been meaning to try it anyway, so I thought I'd give KDE a chance. I've switched to using kdm and letting that log me into either Gnome or KDE at my option, but now I have two other weird problems: under Gnome, I have the command gnome-terminal -e slrn -t slrn --geometry=100x60 --font=gnome-terminal --font=-adobe-courier-medium-r-normal-*-*-120-*-*-m-*-iso8859-1 in my Foot menu to load slrn. This has stopped working from the menu; about 9/10 times it spawns a window that immediately closes, and the other 1 time in 10 it spawns an error box claiming I'm running with glibc 2.1 and I should read my kernel docs because there was an error logging in. Except, it works if I just cutpaste that command into an existing gnome-terminal, and I can run normal gnome-terminals from the Foot menu fine. Meanwhile, under KDE, whenever I log in all my Gnome panels start then promptly start moaning (understandably) that I am not running a Gnome compatable windows manager! I haven't used KDE long enough to determine if my random server death has been cured or not. Normally I enjoy fiddling about with this sort of stuff, but it's starting to get kinda tedious now, particularly the random server death under Gnome. Any suggests for sorting any pieces of this mess out gladly receieved. I'm running testing Ximian, all up to date, with a custom-compiled kernel 2.4.7 (also tried 2.4.5). I'm thinking about trying X 4.1.0 but have stuck with 4.0.3 for the moment in case that just makies things worse. My graphics cards are an AGP Geforce (tried both nVidia and XFree86 nv drvier) and an S3 Virge DX. I'm not sure what other details may be relevent, so I'll hold off and just mail everything anyone asks for rather than swell this mail any more. Thanks in advance! -- -=R=- print join ,reverse split /\s+/,'hacker. Perl another Just',\n; Web: http://www.fscked.co.uk ICQ: 66545073
Re: Problem with apt-get removing a package
On Mon, Jul 23, 2001 at 10:42:56AM -0400, Andrew Dixon wrote: I'd also recommend getting rid of Ximain and installing Gnome from the Debian .debs. It really does eliminate a lot of headaches. Know of any rough instructions on how to do this? Is it just a matter of taking Ximian out of sources.list, apt-get remove *gnome*, then apt-get install *gnome*? -- -=R=- print join ,reverse split /\s+/,'hacker. Perl another Just',\n; Web: http://www.fscked.co.uk ICQ: 66545073
