Looking at the output of netstat, I'm not seeing a definitive way to tell which torrent connections are clients reaching in to my machine vs my machine reaching out to the world. Is there a clear way to determine which is which?
View 1 RepliesIs there a way I can count inbound/outbound mails for a particular user? I'm using sendmail as MTA and dovecot for POP3.
View 1 Replies View RelatedI'm having an issue where a server in CA (1000/full) and in VA (100/full) have very lopsided data transfer.
CA -> VA with iperf shows ~20Mbps
VA -> CA with iperf shows ~93Mbps
If we change the CA server to 100/FULL, transfer speed is 93Mbps both ways.
Some tuning was done to TCP window scaling parameters, but it won't correct the issue, just improve the CA -> VA numbers to what is listed above. I will say, turning TCP window scaling OFF will lower the transfer speed both ways to < 20Mbps.
The only clue I have when looking at wireshark dumps is that the window scale going OUT would never go past 10240 (scale is 8, so 2^8 x 40bytes). In the opposite direction, the window size will go above 3MB (scaled).
It is not a bandwidth problem as iperf with UDP shows 93Mbps both ways. Local transfers (CA 1000/full to CA 100/full) show full speed both ways, so I feel it is strictly related to TCP window scaling.
RedHat 5 64-bit on both sides. Any ideas why it won't scale above 10240?
I am trying to remember the procedure used to determine the ip address range when given a network address and its subnet mask. Does anyone have any documentation on the steps that are done to accomplish this?
View 6 Replies View RelatedI have two NIC cards installed, they both go to seperate networks. Occassionally I am not able to establish a connection via eth1. When looking at tcpdump traces I see that Linux is using eth0 hardware port and putting eth1's IP address in the packet to reply to the connection request. When the connections work its when Linux uses eth1's port.
View 3 Replies View RelatedI have been running into network issues lately with my CentOS 5 (installed on AMD Athlon X2 - 64 bit machine about 2 years ago). It was able to connect to the network fine until yesterday when I installed more memory into the machine restarted it.
Upon restart, the boot up was paused at 'Determining IP information for eth0' step for a while and booted with out connecting to the network. I tried restarting the network (/sbin/service network restart), but this is the error message I continuously get: Determining IP information for eth0...PING xxx.1.226.1 (xxx.1.226.1) from xxx.1.226.83 eth0: 56(84) bytes of data
[Code]...
I've seen this pop up a few times on the forum but my setup is slightly different. I have a PC with an ASUS P5N32-SLI Premium motherboard. I've just installed CentOS 5.1. Yes, I know this is ancient, but this is what they're developing on at work. I'm trying to get connected either via ethernet or wireless and neither one works. When I boot the computer to Window XP it does connect on ethernet. Ethernet should be easy so I'm concentrating on that. But I get "Determining IP information for eth0... failed." after a 'service network start'.Here is some other information that was asked on similar threads:
> uname -rmi
2.6.18-53.el5 x86_64 x86_64
> grep eth0 /etc/modprobe.conf
[code]....
I have installed apf-9.1.7 on linux,after installation inbound activites stopped as per the rules defined.I have added port 20,21 (ftp) in IG_TCP_CPORTS,but still ftp is not working from other Pc.
View 1 Replies View Relatedi wanted to learn linux, so i just installed Centos on a thinkpad x200. my wireless will not work. when i try to activate it i get this error:
Determining IP information for wlan0...SIOCSIFFLAGS: No such file or directory
SIOCSIFFLAGS: No such file or directory
# lspci
00:00.0 Host bridge: Intel Corporation Mobile 4 Series Chipset Memory Controller Hub (rev 07)
[code]....
I'm having a problem that seems to plague a lot of people judging from my research on the web. I have a hosting provider that limits the number of incoming connections to the shared host to 50 per IP.
I have a single IP for outbound connections and I use Squid as a proxy server.
Lately I've tripped across the 50 connection limit frequently - and that's with only 1 user. It seems the problem is related to the performance you can get out of a desktop these days. Its not impossible to have several browsers open with several connections to different sites on the same server - and boom - locked out!
So it occurred to me that there must be some way to limit the number of outbound connections in the kernel - but I've not found it. I did find that Microsoft had been limiting the number of outbound connections in XP to 10 to address the virus problem, and I've found countless hosting complaints and dialog on the subject with no easy solution.
So my question is simply, does anyone know how to limit the number of OUTBOUND connections to a single IP in the kernel?
Is there a way to configure my interface to promisc mode and also make it not capture the "transmitted" packets. ?I mean, i want the interface in Promisc mode but only for inbound traffic.If there isnt any using ifconfig, can it be by configuring eth0 to promisc using ifconfig , and filtering outbound traffic from being captured using sockets or something ?
View 4 Replies View RelatedI have a system running 11.04 and it is dropping packets on the hardwired ethernet interface to other systems on the LAN, only in the inbound direction. It drops packets every 5 seconds. I verified this with iperf. Outbound packets pass with no problems. The network card in this system is a Broadcom Corporation NetXtreme BCM5752
View 9 Replies View RelatedI have problem on VPS running opensuse. When I enable firewall outbound connections stop working. I have tried everything I know (not much when it comes to firewall (iptables)) but could not solve this.
Here is my ifconfig:
Code:
I used xxx.xxx.xxx.xxx to hide real address.
I've read up some of the posts on this forum, but can't seem to find an answer. I have a web service within an Apache Tomcat instance installed on a Redhat linux server. I only have shell access to the server, and need to monitor outbound network traffic from my web service. Is there a unix command that will allow me to monitor all outbound traffic? I'm thinking fiddler, but a unix version? I've heard of things like ntop and iptraf, but I don't think those will help me in this instance.
View 2 Replies View RelatedI'm running a fresh install on an intel i7 system on an asus p6t deluxe v2 motherboard with the onboard NIC (Marvell Technology pci-e). I know the NIC is working as when I boot into the onboard OS that Asus provides (Asus Express Gate SSD)I have a working net connection.Booting up CentOS hangs at determining IP information for eth0 and eventually fails with the following error:PING xxx.170.30.1 from xxx.170.31.231 eth0: 56(84) bytes of data---xxx.170.30.1 ping statistics ---4 packets transmitted, o received, +3 errors, 100% packet loss, time 2999ms, pipe 3 failed.
RTNETLINK answers: file exists.
RTNETLINK answers: file exists.
RTNETLINK answers: file exists.
[code]....
I have a script I only want to run when a window is being dragged around. Is there an environmental variable or something that will tell me whether one is moving and which one it is? I'm using Ubuntu Lucid if it matters.
View 9 Replies View RelatedHow to check inbound and files transfer from outbound to inbound in my linux server. because in my server from outside to inside transfer is high. so its take huge bandwidth. how to check this and reduse this .
View 2 Replies View Related1. It is definitely my system and not the router. I had been accepting these connections just fine on my Fedora 6 system, but upgraded to FC11 and haven't been able to get these inbound connections working. I still have the drive with the FC6 system and whenever I put the old drive in the system, the connections work fine. So, the hardware, wiring, etc is identical between the FC6 and FC11 systems.
2. The system's firewall is disabled. iptables, and ip6tables are turned off both at startup (e.g., chkconfig iptables off) and using "service iptables stop". (Note--my router also works as a firewall)
3. sshd and httpd are both on and listening on all interfaces. "nmap -P0 -p 22,80 127.0.0.1" shows both ports open, as does 192.168.1.80. But when I use the same command looping through the external IP, the ports are filtered:
Host is up.
PORT STATE SERVICE
22/tcp filtered ssh
80/tcp filtered http
4. hosts.deny is empty.
5. The messages and secure logs do not show anything in response to the nmap scans.
Again, I'd have sworn this was my router, except that it cleanly allows the traffic when I swap the drives in my system.
I'm facing a strange problem. I have a Debian squeeze machine connected to Internet through a 3G USB modem. The machine connects as expected, and I can resolve domain names and establish HTTP and SSH connections.The problem is when I try to communicate to that machine from another PC connected to Internet. The machine doesn't respond to PING and I can't connect to the SSHD installed in it. If I connect the machine to the local network, it works right.
route:
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
default * 0.0.0.0 U 0 0 0 ppp0[code]...........
I am running Ubuntu 10.10 I have an question about the firewall Firestarter, when checking the firewall it told me there are 9 serious incoming connections what must I do with this info. Inbound is normally blocked as standard i have also see that someone with port 1234 and 12345 have trying to attempt mine system but failed all trojan ports are fully blocked.
View 2 Replies View RelatedI'm trying to adjust the firewall to only inbound syn connections.
To Allow all home subnets access to port 53 both tcp/udp but deny the rest.
Using a Linux program to fix my TiVo. Program is burned to a bootable CD, which loads fine. I get asked what the drives are (CD and Hard Disk) with abbreviations like hda, sda, etc. I was able to determine the Hard Drive as /dev/sda, by using <Shift>-PgUp/PgDn, but could not find what the CD drive was.
View 1 Replies View RelatedI have and embedded system which apparently, due to space considerations, is missing lots of normal commands. I can get the following information by "cat /proc/version":
Linux version 2.4.19-uc0 (...) (gcc version 2.95.3 20010315 (release)(ColdFire patches - 20010318 from http://fiddes.net/coldfire/)(-msep-data patches)) #461 Tue May 24 12:06:06 PDT 2011
At issue is the fact that, after some time, the system stops responding to the Network (e.g., you can ping it, then some time later you can't). We suspect that some sort of network monitor is accessing the device and causing the trouble, but don't know for sure.
What I would like to do is log some sort of useful network status information and/or be able to reset the network if I can figure out there is an issue from within the device.
I thought I had it all working using "/etc/init.d/network restart" but I've come to find out that the command is not recognised in this system. I also can't issue a "top" command to see what services are running. I can Telnet in, so I can try commands and see if they work.
Have any of you worked with such an abbr. system and if so, what might be around that I'm unaware of to help determine network status?
This questions pertains mainly to building a brand new GNU toolchain on one machine for deployment on fs of another machine that has no toolchain installed. I understand the machine on which GNU toolchain is going to be built as the 'build' machine. I understand the machine on which GNU toolchain is going to be deployed as the 'host' machine. I understand the machine on which the code generated by the new GNU toolchain is going to be run as the 'target' machine.
Q1.) Is there a fixed 'rule', 'matrix' or 'table' somewhere which tells me exactly which versions of 'binutils' and 'libc' are required by a particular version of GCC? I.e. something along the lines of:
- gcc ver.'k' can be built using gcc ver.'m'(or later) with binutils ver.'i' and libc ver. 'j'
- gcc ver.'k' will operate with binutils ver.'t' and libc ver.'r'(or later)
Is there some 'rule of thumb' / 'heuristics' giving me the above version requirements for the building and operating GCC, or is this information somehow implied or derived systematically. Where can I find more about the above?
Q2) I assume that the libraries have no dependencies on any underlying software other than the system calls provided by the kernel. Furthermore it appears that the libraries are not too sensitive about the kernel version they run on (I found this out from experience). Therefore I could build the libraries on some build machine and place them on the bare naked newly set up host file system without regard to any other prerequisite SW components?
Q3) During operation, do GCC and its various components place system calls directly to the kernel or via library calls (or both)? In other words I am trying to understand conceptually if GCC requires 'Glibc' for its own operation or not?
Q4) During operation, do the 'binutils' components place system calls directly to the kernel or via library calls (or both)? In other words I am trying to understand conceptually if the 'binutils' require 'Glibc' for their operation or not? ( Of course it is clear to me that when GCC is operating it will need libraries and headers to resolve object references in the target code it produces, but that would be a >different< question).
I am trying to scroll text from a text file across a wx.panel. This is the code so far:
Code:
def scrolling(self, text, rate):
def scroll(): charList = []
dc = wx.ClientDC(self.panel_1)
open_text = open(text, "rb")
my_text = open_text.read()
charList.append(my_text) open_text.close()
x,y = self.panel_1.GetClientSizeTuple()
while 1: for z in charList: for r in range(2000):
dc.DrawText(z,x-r,0) time.sleep(rate) dc.Clear()
net3 = threading.Thread(target=scroll)
net3.setDaemon(1) net3.start()
It scrolls but I am guessing where the range is (2000). I need to be able to determine the total number of pixels that the text in the file is. Is there a python way to determine this?
We are planning on rolling out a mail service that will backend a webmail/groupware interface using postfix and dovecot, and we are trying to nail down an architecture for the system. I was hoping I could get some opinions from you good people on the different models we have come up with, and any improvements. To preface this discussion, we use NFS as the backend storage for mail so that clusters of mail servers can be created.
One possible solution is to separate three services, inbound email, outbound email, and message retrieval onto separate servers. We currently run a mail system with 65K users in this manner and it works, but there is a lot of wasted cpu on the inbound and outbound mail servers. where the dovecot servers are doing most of the work.
Another solution would be to combine the dovecot and inbound postfix servers to run dovecot as the local transport agent for postfix with LMTP, giving us quota and sieve capability.
The last solution is to run all three services on the same machine, giving us a single server image that we can clone on demand when we need to add more performance.
The single server, all three services, solution seems to be the simplest to build and manage, but putting all your eggs in one basket like that concerns us. What if there is a sudden influx of mail from our filtering platform, and it slows down all the other processes? What if the NFS server becomes disconnected from the network and the load average on each server skyrockets killing the ability for outbound mail to flow (where a separate outbound mail server would continue running despite this issue)?
RHEL 5.4 i want to be able to do redistribution of inbound syslog messages to syslogd. as example, my syslog.conf has in it at the end:
*.* @192.168.5.5
*.* @192.168.5.6
my sysconfig/syslog file has "-r" as the only option for syslogd. any messages generated by the localhost will be sent to the two remote servers, but messages that come into this box (udp 514) only get logged locally and do not get sent out to the remote hosts.
you may ask why do i want to do this. because i have several syslog servers (for security purposes) and many of my net devices are configured to send syslog to all the syslog servers, hence each device is sending way too much duplicate udp-514. so i would like to minimize the udp-514 coming out of the devices, have all devices send to a central syslog server, and then central syslog server do distribution to the other syslog servers. others have also called this "syslog proxy". or, if not with syslogd, how to achieve this (preserving the original syslog message host info, etc)?
We host a web server in which we are hoping to implement some form of traffic redirection based on source IP address, and I am wondering whether the squid proxy built on iptables would be capable of managing this task? Essentially we are trying to redirect traffic from specific set of source IP ranges to a "Your IP has been restricted" type of page at a different IP/FQDN.
View 2 Replies View RelatedI need to log outbound (server -> remote) ssh traffic, but am unsure of the best way to do this.
I added this to iptables, but nothing is being written to the logs when I ssh from the server to another system.
iptables -A OUTPUT -m state --state NEW -j LOG --log-uid iptables-save
I have internal-only email server that has internal BIND9 running. Thought it only has its IP address defined in /etc/resolv.conf, it is still resolving outside addresses.
View 2 Replies View Related