syslogd: select: Bad file descriptor
Feb 10 08:07:45 xx syslogd: select: Bad file descriptor FreeBSD 4.9-RELEASE-p1 i386 run as in a ttys wrapper script: /usr/sbin/syslogd -4 -A -cc -n -s -d -f /etc/syslog.conf the output of script/command is not redirected to /var/log but i suspect it happens when newsyslog runs. Has anyone encountered the same? Does anyone know of a remedy? _// Sten Daniel Sørsdal ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
PPP and radius.conf - shouldnt it be doing round-robin?
I use userland ppp with radius authentication against 2 radius servers, on a freebsd 4.8-stable box that i use as a router/gateway. The two servers are on two different interfaces, it seems that when i unplug the first server (#1 in radius.conf) from the switch the arp cache on the gateway will time out and a "Host is down" message is generated (or icmp host unreachable). All authentication requests are then automatically denied without consulting the second server. If i then swap the order of the radius servers in radius.conf while the first server is still down, i get authenticated. If i arp -s the mac address of the unplugged server, and it's the first one in radius.conf, the authentication mechanism proceeds to query the second server and i get authenticated. Is this intended or is it one of those icmp unreach/host down issues i've seen more and more often lately? i read this in 'man radius.conf': Up to 10 RADIUS servers may be specified for each service type. The servers are tried in round-robin fashion, until a valid response is received or the maximum number of tries has been reached for all servers. uname -a: FreeBSD fictious 4.8-RELEASE FreeBSD 4.8-RELEASE #0: Sun Aug 3 00:55:37 GMT 2003 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/WACCESS i386 - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Syslogd dying often on different hardware - hackers?
Many of my servers tend to have their syslogd die on them. In dmesg i could see the signal number on one of them; 11 But kill -l doesnt reveal which that is: pid 17263 (syslogd), uid 0: exited on signal 11 pid 87264 (syslogd), uid 0: exited on signal 11 /var/log has lots of space, newsyslog does it's job regularly. The versions vary from 4.7 to 4.9-prerelease. Lots of memory available. Hardware ranges from 1ghz celerons, 400mhz p3's, 1.6ghz p4's. The command lines are either; -nccvvs -ccvvs (but mostly) -vvs They log to remote host that is currently unavailable. (router emites destination host unreachable). The reason i mention this is that many types of software misbehave when they receive this message. When i ran syslogd -dvvs in 'screen' it started logging to remote 'loghost'. For a -long- time it kept repeating "Host is down" and tried to log to 'loghost' 'loghost' is a local alias. I'm finding it hard to reproduce. -- Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Dumping/reading memory space of a running program.
How do i read/dump the entire memory space of a running program? Is this possible? I really dont have the option of restarting it. -- Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Requirements for doing a spin-off bsd branch?
I am curious as to what is required (legally) to make a branch of FreeBSD and call it somethingelseBSD? This would be non-commercial but is it possible to do a commercial version as well? Any thoughts, advice and opinions are appreciated. - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: How do I max a 6Mbps link
> > Hi all, > > What configuration changes do I need to make to two > freebsd-stable boxes to > fully max out a 6Mbps/220ms network link? This is for bulk 500+MB file > transfers. > > The target application is proftpd with ncftpd as the client. > > Thanks in advance, > Max > 220ms? I dont think TCP can handle this. Look for a non-connection oriented protocol to transfer files. UDP for example, or better, raw IP. Maybe you're lucky and get FAST to work :-) - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: Changed ISP now can't get to websites / traceroute
> The thing is, > now that I've got this new ISP (Zen - 8 static IP's), I find > that there > are sites that I am not able to get (browse) to, most concerning is > www.freebsdforums.org. Ever since I've got this new ADSL account, not > for hell nor high-water can I hit that site, using any browser, using > any OS (FreeBSD, RH £inux or MS Win2K Pro / XP) at home. There could be a hundred reasons as to why you dont reach www.freebsdforums.org. _Might_ be an MTU problem somewhere upstream, if ICMP is filtered this is definitly a bigger problem. Does ping work? If they also filter UDP as you mention (fbsd traceroute not working) I'd suggest you talk to your ISP and find out exactly what they are filtering. Are those 8 static IP's public or private? If they start with 192.168.x.x, 10.x.x.x, 172.16.x.x through 172.16.31.x.x they must be NATed and it is their NAT configuration that blocks the traceroute. Contact the ISP! - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
nano / nanorc
FreeBSD 4.8 Im trying to customize nanorc but i cant seem to either get the syntax activated or the colors to work (hard to tell). Does anyone know what the requirements are for using color? Oh, im doing this remotely via ssh using securecrt. Nano v 1.2.1 built from ports. - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Running out of ttyp's on 4.8 RELEASE
I'm having trouble with running out of ttyp's on my 4.8 Release box because of "screen". When I do at 'who' the ttyp list stops at ttypv. This makes it impossible to log in to the box with telnet or ssh. I've increased pty's as suggested in the archives but that doesn't make any difference. How do I increase the ttyp's? - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
IPFW stateful deny question
This one is for the archives. If anyone would donate their time to replying to this one, It would be appreciated. Im trying to setup a firewall that for a time-limited period block IP's that send packets to specified ports. After time has expired the ip will be allowed again. With this setup; ipfw add 100 check-state ipfw add 1000 skipto 2000 tcp from any to any 445 ... ipfw add 2000 deny ip from any to any keep-state ... Would this setup a stateful block against that IP address, blocking all IP traffic from that particularly evil ip Or would it just match the specifics? Would I have to do this instead to achieve this? ipfw add 100 check-state ipfw add 1000 skipto 2000 tcp from any to any 445 keep-state ... ipfw add 2000 deny ip from any to any ... - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: Dual homed host routing problem
> I'm running FreeBSD 4.7-RELEASE and I have trouble routing > between two NIC's. On one side I have a 192.168.1.0/24 > network and on the other a 212.110.94.64/27 network on which > I have mail and web servers, which the 192.168.1.0/24 hosts > should be able to reach. > > Here are the ifconfig and netstat -r outputs: > wb0: flags=8843 mtu 1500 > inet 212.110.94.84 netmask 0xffe0 broadcast 212.110.94.95 > inet6 fe80::280:48ff:feb5:af3%wb0 prefixlen 64 scopeid 0x1 > ether 00:80:48:b5:0a:f3 > media: Ethernet autoselect (100baseTX ) > status: active > rl0: flags=8843 mtu 1500 > inet 192.168.1.1 netmask 0xff00 broadcast 192.168.1.255 > inet6 fe80::202:44ff:fe4f:958e%rl0 prefixlen 64 scopeid 0x2 > ether 00:02:44:4f:95:8e > media: Ethernet autoselect (10baseT/UTP) > status: active > > Routing tables > Internet: > DestinationGatewayFlagsRefs Use > Netif Expire > default 212.110.94.65 UGSc40wb0 > localhost localhost UH 00lo0 > 192.168.1 link#2 UC 1 > 0 rl0 > 192.168.1.255 ff:ff:ff:ff:ff:ff UHLWb 1 45rl0 > 212.110.94.64/27 link#1 UC 80wb0 > > and I have net.inet.ip.forwarding set to 1 > > How do I get my box to route packets between the two > interfaces 192.168.1.1 and 212.110.94.84? Log into 212.110.94.65 and tell it that 192.168.1.0/24 is behind 212.110.94.84. If it's a BSD box you could do on .94.65; route add -net 192.168.1.0/24 212.110.94.84 - Sten ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: Equal-cost multipath routing for FreeBSD 4-STABLE?
It is unclear wether one2many support bundling through VLANs as well. Any clues? - Sten -Original Message- From: Matthew Seaman [mailto:[EMAIL PROTECTED] Sent: 20. mars 2003 14:25 To: [EMAIL PROTECTED] Subject: Re: Equal-cost multipath routing for FreeBSD 4-STABLE? On Thu, Mar 20, 2003 at 03:15:27PM +0200, Andrew Stesin wrote: > suppose we have FreeBSD box `A' with 3 network interfaces (ip > unnumbered or numbered - shouldn't matter). This 3 interfaces form 3 > parallel IP channels to some single destination `C' (let's think of > another FreeBSD box `B' with another 3 interfaces at the other end and > `C' being an alias to lo0 of `B' box). > > Than let's consider either 3 equal static routes to `C' set at `A', or > OSPF with 3 equal cost routes to `C' from `A'. > > Is it possible to tell kernel to route packets between `A' and `B' in > a round-robin way, so that each of 3 channels handle 1/3 of a total > consumed bandwidth? Something like "equal-cost multipath" feature of > Cisco IOS? Yes. See the ng_one2many(4) man page. You don't need to futz about with aliases to lo0 on machine C: just make another ng_one2many cluster there and treat both of them as a regular network interfaces. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. 26 The Paddocks Savill Way PGP: http://www.infracaninophile.co.uk/pgpkey Marlow Tel: +44 1628 476614 Bucks., SL7 1TH UK To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
TCP in TIME_WAIT for too long.
Im using a 4.7-STABLE box with Zebra BGPD for full BGP feed and i have come across an unusual problem. It works great except when i come up in the situation when the provider reboots their router. FBSD doesnt seem to detect that the link has been broken and the TCP connection will stay in TIME_WAIT for a long time (up to 20 minutes?). Is there any way to tweak this to die after about 120 seconds? Will this break anything? Any help is appreciated! - Sten To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
RE: PPPoE and troubles with TCP
Have you remembered to setup anything to fix the MSS in TCP packets going through? You have disabled encryption and compression right? -Original Message- From: Rostislav Krasny [mailto:[EMAIL PROTECTED]] Sent: 26. desember 2002 15:42 To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED]; FreeBSD Questions Subject: Re: PPPoE and troubles with TCP --- Stacey Roberts <[EMAIL PROTECTED]> wrote: > On Thu, 2002-12-26 at 13:16, Rostislav Krasny wrote: > > --- Stacey Roberts <[EMAIL PROTECTED]> wrote: > > > On Thu, 2002-12-26 at 12:26, Rostislav Krasny wrote: > > > > Hello > > > > > > > > I have Alcatel Speedtouch Home ADSL modem. I use PPPoE protocol > > > > for connection with my Internet Service Provider (ISP) through > > > > this > ADSL > > > > modem. I use FreeBSD 4.7-RELEASE and ppp (a.k.a. user-ppp) for > this. In > > > > the /etc/ppp/ppp.conf file I have ' set MRU 1492' and ' set MTU > 1492' > > > > options. This is because the maximum MTU one can use with PPPoE > > > > is > 1492 > > > > according to RFC 2516. I have no problems with "dialing" to my > > > > ISP. > But > > > > after the PPPoE connection is established I have troubles with > > > > TCP connections to some hosts. For example I can't go to > > > > hhtp://www.ssh.com/ but to http://www.freebsd.org/ or > > > > http://www.gnu.org/ I can. If I run 'telnet www.ssh.com 80' and > enter > > > > HTTP request according to HTTP 1.1 or 1.0 protocol I don't get > > > > any answer. If I enter just 'GET /' I get some short answer. > > > > 'GET /' is > an > > > > HTTP request according to the old HTTP 0.9 version of the > > > > protocol > and > > > > it is rarely used today. But the problem is not HTTP, I think. > > > > If I > do > > > > HTTP request according to HTTP 1.1 from my friends' computer I > > > > do > have > > > > some long answer. Furthermore if I change the MRU and MTU to > > > > 1484 > or > > > > smaller I have no problem with www.ssh.com. So it looks like TCP > packet > > > > overflow or something like that. > > > > Besides FreeBSD I have Windows 98 Second Edition in my computer. > > > > I > use > > > > PPPoE with Win98SE too. I use RASPPPOE driver fo it, go to > > > > http://www.raspppoe.com/ for more information. I have no > > > > problems > with > > > > www.ssh.com when I use Win98SE with this impimentation of PPPoE > > > > protocol. I have MTU == 1492 set by default in RASPPPOE. I don't > get > > > > error messages about fragmentation when I run 'ping -f -l 1464 > > > > www.ssh.com' in my Win98SE (ICMP+IP=1464+28=1492). Why in > > > > FreeBSD I cannot use MTU == 1492 without problems with some > > > > hosts like www.ssh.com ? How to fix it in FreeBSD? > > > > > > > > P.S. ping have different syntax in Windows, so don't be wondered > about > > > > -f flag. -f sets "no fragmentation" flag in IP headers of all > outgoing > > > > packets. > > > > > > I use ADSL here on my site, but not that modem. How do you connect > > > to that device from the FBSD box? > > > > I configured my ADSL modem to work with PPPoE. Then I changed the > > /etc/ppp/ppp.conf file to this one: > > > > /etc/ppp/ppp.conf > > default: > > set log Phase Chat LCP IPCP CCP tun command > > ident user-ppp VERSION (built COMPILATIONDATE) > > > > set device PPPoE:ed0 > > set speed sync > > # set MRU 1484 > > # set MTU 1484 > > set MRU maximum 1492 > > set MTU maximum 1492 > > set timeout 0 > > set urgent none# no priorities > > enable dns # request DNS info (for > > resolv.conf) > > disable ipv6 ipv6cp > > disable shortseq protocomp acfcomp vjcomp deflate pred1 > > mppe > > denyshortseq protocomp acfcomp vjcomp deflate pred1 > > mppe > > > > myisp: > > set authnameusername@ISPstring > > set authkey password > > add default HISADDR # Add a (sticky) default route > > /etc/ppp/ppp.conf > > > > The hostname of my box can be resolved through /ets/hosts , so I > > don't need any kind of 'set ifaddr' in /etc/ppp/ppp.conf file. ed0 > > is the ethernet NIC my ADSL modem is connected to. > > > > Many lines of the "default:" section of the ppp.conf file are not > critical. > > The only ones you need are these: > > > > set device PPPoE:ed0 > > set MRU maximum 1492 > > set MTU maximum 1492 > > enable dns > > > > See `man 8 ppp` for more information. > > > > Dude, > I wasn't asking for assistance.., rather I was hoping to provide > some help in response to your post.. :-) > > The reason why I asked about how you connect to your ADSL modem is > because if its ethernet, then there is no need to use ppp. > > But if there's a case-requirement why you need to use ppp, then fair > enough. > > Let me know e
RE: duplicate packets in ping?
I've had similar issue with Intel Ethernet controllers (i82562ET and i82801BA/BAM). It was practically no load, and on the wire there was no duplicates but ping got duplicates anyway. - It only happens sporadically and it happens to about 10 boxes with the same FreeBSD version (exact same). Havent had the chance to dig into it though Same with you? -Original Message- From: Randall Stewart [mailto:[EMAIL PROTECTED]] Sent: 20. november 2002 13:34 To: shubha mr Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: duplicate packets in ping? shubha mr wrote: > Ping on my device driver (for my NIC )gives duplicate packets.does > anyone know why they occur and how to eliminate them? > thanks > shubha > > __ > Do You Yahoo!? > Everything you'll ever need on one web page > from News and Sport to Email and Music Charts http://uk.my.yahoo.com > > To Unsubscribe: send mail to [EMAIL PROTECTED] > with "unsubscribe freebsd-net" in the body of the message > > I have noticed a similar occurance in some of my SCTP testing with all of the BSD's. In particular the linksys pcmcia cards seem to do this. In my testing and analysis this always seems to occur when the card is busy and what happens is you lose some packet and another appears to be duplicated... I traced this out with ethereal a while ago and then just stopped using that card when I figured out it was some sort of circular buffer issue.. I did not dig in and find out if it was the card or driver... I rather suspect it is the card (since it is a low end one).. but one never knows... R -- Randall R. Stewart [EMAIL PROTECTED] 815-342-5222 (cell phone) To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-net" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
