i am trying to run this script , and it shows no tcp connections any ideas? OS: centos 6.3
#! /usr/bin/env stap
probe begin {
printf("%6s %16s %6s %6s %16s\n",
"UID", "CMD", "PID", "PORT", "IP_SOURCE")
}
probe kernel.function("tcp_accept").return?,
kernel.function("inet_csk_accept").return? {
sock = $return
if (sock != 0)
printf("%6d %16s %6d %6d %16s\n", uid(), execname(), pid(),
inet_get_local_port(sock), inet_get_ip_source(sock))
}
OUTPUT:
[root#server src]# ./tcp_con.stp
Missing separate debuginfos, use: debuginfo-install kernel-2.6.32-279.1.1.el6.centos.plus.x86_64
UID CMD PID PORT IP_SOURCE
Follow systemtap's advice and install kernel-debuginfo.
The question marks after both kernel.function() probes let stap quietly drop both of those.
Related
I am comfortable with writing single queries. I am new to writing bash scripts trying to automate the daily stuff. I need help on creating a PowerShell or bash script where I can SSH to multiple Linux devices with same SSH key and then reboot the devices.
I access linux devices manually with the following command in PowerShell
ssh -i C:\<path-to-the-private-key\test.ppk test#XX.X.X.XXX (X - IP Address)
Then I enter the following command
sudo reboot
It asks me to type the password and then restarts the device.
I have 100+ devices that I need to restart.
I can get a list of all IP address in a text file. How can we search for all the IP address in a text file, authenticate with the SSH private key and run the sudo command to restart the device?
Would it also be possible to throw a message if it was not able to restart a device?
Any help would be appreciated.
This is the script that I have.
testreboot.sh
#!/bin/bash
pw="test123"
hosts='IP.txt'
while read -r line; do {
/usr/bin/expect << EOF do
ssh test#"$hosts" 'sudo reboot'
expect "*?assword*"
send "%pw\r"
EOF
}
done < $hosts
IP.txt
XXX.XX.XX.XX
XXX.XX.XX.XX
XXX.XX.XX.XX
XXX.XX.XX.XX
I have Ubuntu 20.04 installed from Windows App Store. I am trying to run the testreboot.sh from PowerShell using the following command and get the following error message.
bash testreboot.sh
testreboot.sh: line 2: $'\r': command not found
testreboot.sh: line 3: $'\r': command not found
testreboot.sh: line 5: $'\r': command not found
testreboot.sh: line 16: syntax error near unexpected token `done'
testreboot.sh: line 16: `done < $hosts'
A better solution to this problem is to use something like Ansible, Chef, or Puppet to solve these multi-server coordination needs.
Here is an example in a shell script using expect to ssh to another server and login:
NOTE: When you use expect in this manner, you need to escape " and $ and
other items. If password has $, it must be escaped.
This is where after logging in and expect see a command prompt
For example:
44 [localhost.localdomain]/home/map%
This is where you would need to add sudo reboot command
-re \"$reg_ex_prompt\" {
}
Test Script:
#!/bin/sh
#
debug=0
exit_val=0
spawn_item="ssh"
destination="user_name#<IP of server to shh>"
reg_ex_prompt='\[%|>|\$|#\] $'
#
# Change -D 0 to -D 1 to debug interactivly
#
expect -D $debug -c "
spawn $spawn_item $destination
set ret 1
set timeout 20
expect {
timeout {
send_user \"Timeout reached!\"
exit \$ret
}
eof {
puts \"End of test connection reached!\"
if { \$ret == 0 } {
puts \"Connection test Successful!\"
puts \"Exiting $destination ...\"
} else {
puts \"Connection Failure!\"
}
exit \$ret
}
\"Connection refused\" {
puts \"ERROR: Trouble connecting to $device_type_parm destination $destination\"
puts \"Aborting...\"
exit \$ret
}
\"Permission denied\" {
puts \"ERROR: User name or password is incorrect for $destination\"
puts \"Aborting...\"
exit \$ret
}
\"Connection reset by peer\" {
puts \"ERROR: Trouble connecting to $destination\"
puts \"Aborting...\"
exit \$ret
}
\"you sure you want to continue connecting\" {
send \"yes\r\"
exp_continue
}
\"assword:\" {
puts \"\r\nSending password\"
send \"password\\\$\r\"
exp_continue
}
\"Choice? \" {
send \"1\r\"
exp_continue
}
\"Q. Quit\" {
send \"q\r\"
exp_continue
}
-re \"$reg_ex_prompt\" {
send \"sudo reboot\r\"
sleep 2
set ret 0
exit \$ret
}
interact
} "
# get the exit value from expect statment above
exit_val=$?
As I know, tcp port in 'time_wait' stat cann't be used. However, in my experiment, server reuses the 'time_wait' port? Why?
Firstly, in client machine, type command ehco 40000 40001 > /proc/sys/net/ipv4/ip_local_port_range. So, the maximum number of TCP ports is 2.
server code
while (1) {
int len = sizeof(struct sockaddr);
fd = accept(sfd, &remote, &len);
read(fd, buf, sizeof(buf));
close(fd);
}
client code
for (i = 0; i < 3; i++)
{
sleep(1);
pid_t pid = fork();
if (pid == 0)
{
handler();
exit(0);
}
}
void handler()
{
* ............. */
res = connect(sfd, result->ai_addr, result->ai_addrlen);
if (res == -1) {
perror("error");
exit(1);
}
printf("connect\n");
}
show
[root#livecd ~]# ./client
connect
[root#livecd ~]# connect
connect
It's up to 3 connections. I think, 2 connections at most. Why ?
server has 2 timewait connections.
[root#livecd ~]# netstat -anp | grep TIME
tcp 192.168.88.131:2016 192.168.88.132:40000 TIME_WAIT
tcp 192.168.88.131:2016 192.168.88.132:40001 TIME_WAIT
Environment
Linux livecd.centos 2.6.32-642.el6.i686 #1 SMP Tue May 10 16:13:51 UTC 2016
server config
[root#livecd ~]# cat /proc/sys/net/ipv4/tcp_fin_timeout
60
[root#livecd ~]# cat /proc/sys/net/ipv4/tcp_tw_recycle
0
[root#livecd ~]# cat /proc/sys/net/ipv4/tcp_tw_reuse
0
client config
[root#livecd ~]# cat /proc/sys/net/ipv4/ip_local_port_range
40000 40001
Important
I also try ubuntu server 14.04, but got the same result.
I am trying to print a file in remote server through CUPS command , User needs to know the status of job status. How to get a response from that. Here is my code :
#!/usr/bin/perl
my $response = system("lpr -P laserJet123 -o raw -T test_womargin abc.txt");
print $response;
$? returns the response status of system command.
system("lpr -P laserJet123 -o raw -T test_womargin abc.txt");
if ($? == -1) {
print "failed to execute: $!\n";
}
elsif ($? & 127) {
printf "Job failed with signal %d, %s coredump\n",
($? & 127), ($? & 128) ? 'with' : 'without';
}
else {
printf "Job exited with value %d\n", $? >> 8;
}
I have perl script
perl script (install.pl):
use strict; use warnings;
use Term::ANSIColor;
my $logfilename = "install.log";
open LOG,">$logfilename" or die "failed to open log file reason:$!";
sub logMsg
{
my ($msg) = #_;
my ($error) = $_[1];
$msg .= "\n";
print LOG $msg;
if( $error ) { print color 'red' ;}
{print $msg}
if( $error ) { print color 'reset' ;}
}
sub lampInstall
{
logMsg("Installing apache2, php, and mysql db");
# Install apache2 and mysql:
my $cmdOutput = `./ubuntu-lamp-install.sh`;
print LOG "$cmdOutput\n";
}
lampInstall();
close LOG;
bash script (ubuntu-lamp-install.sh)
#!/bin/bash
sudo apt-get update
sudo apt-get install ntp
sudo /etc/init.d/ntp start
export DEBIAN_FRONTEND=noninteractive
echo "mysql-server mysql-server/root_password password test" | sudo debconf-set-selections
echo "mysql-server mysql-server/root_password_again password test" | sudo debconf-set-selections
sudo apt-get install -y apache2 php5 libapache2-mod-php5
sudo service apache2 restart
sudo a2enmod ssl
The issues is that , when install.pl is invoked, it waits for long time and does not give any output information of what is being installed.I need it to change the perl script to display the output information from bash script(better if instantaneous) as well as log the output to log file. The script is written by someone else , I have very little knowledge of perl scripting.
You could do something like
open my $cmdOutut, "-|", "./ubuntu-lamp-install.sh";
while (<$cmdOutput>) {
print LOG "$_\n";
print "$_\n";
}
in your lampInstall().
This is the new lampInstall():
sub lampInstall
{
logMsg("Installing apache2, php, and mysql db");
# Install apache2 and mysql:
open my $cmdOutut, "-|", "./ubuntu-lamp-install.sh";
while (<$cmdOutput>) {
print LOG "$_\n";
print "$_\n";
}
}
I built a Linux kernel with CONFIG_IP_MULTICAST=y,however no UDP multicast package received in this kernel while UDP unicast works well.
ethtool -S eth0 | grep multicast
txmulticastframes_g: 0
txmulticastframes_gb: 0
rxmulticastframes_g: 0
Any hints how can I solve this problem?
Thx. Forrest G
=================================================================================
Additional:
tcpdump can get the packet
root#JHI # ./tcpdump port 3702
device eth0 entered promiscuous mode
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on eth0, link-type EN10MB (Ethernet), capture size 65535 bytes
-7:-48:-19.4151 IP 192.168.42.212.3702 > 239.255.255.250.3702: UDP, length 787
-7:-48:-19.4661 IP 192.168.42.212.3702 > 239.255.255.250.3702: UDP, length 803
^C
2 packets captured
2 padevice eth0 left promiscuous mode
ckets received by filter
0 packets dropped by kernel
I write this WS-devicediscovery function with gSOAP. It works on X86 machine. When running on ARM device, it can send out igmp packet but not receive anything.
void wsdd()
{
struct soap *soap_udp;
struct ip_mreq mreq;
soap_udp=soap_new();
soap_init1(soap_udp, SOAP_IO_UDP|SOAP_IO_FLUSH);
if (!soap_valid_socket(soap_bind(soap_udp, NULL, 3702, 100)))
{
soap_print_fault(soap_udp, stderr);
}
mreq.imr_multiaddr.s_addr = inet_addr("239.255.255.250");
mreq.imr_interface.s_addr = htonl(INADDR_ANY);
if(setsockopt(soap_ud->master,IPPROTO_IP,IP_ADD_MEMBERSHIP,&mreq,sizeof(mreq))==-1) {
perror("membership error\n");
}
int loop = 1;
int sock_opt = 1;
if ((setsockopt(soap_udp->master, SOL_SOCKET, SO_REUSEADDR, (void *) &sock_opt,
sizeof (sock_opt))) == -1) {
printf("setsockopt\n");
}
if ((setsockopt(soap_udp->master, IPPROTO_IP, IP_MULTICAST_LOOP,
&loop, sizeof (loop))) == -1) {
printf("setsockopt\n");
}
while(1){
soap_accept(soap_udp);
soap_serve(soap_udp);
soap_end(soap_udp);
}
}
Then I have tried these things but still not work
route add -net 224.0.0.0 netmask 240.0.0.0 eth0
echo 0 > /proc/sys/net/ipv4/conf/all/rp_filter
Problem solved by changing..
ifconfig eth0 promisc
Can anyone explain the principle?