Eytan's questions

I installed a vpnc on my Linux red-hat machine version 6.3

Most of the cases vpnc work well ( connecting to gateway customers by vpnc )

But against some particular customers ( Cisco gateways ) , immediately after I activate the vpnc I get the following messages:

    vpnc: quick mode response rejected:  (ISAKMP_N_INVALID_PAYLOAD_TYPE)(1)

In this case Vpnc could not be connected

Please advice what the problem here? ,

 [root@TEST-LINUX-MACHINE /var/tmp]# vpnc
 Enter password for [email protected]: 


Connect Banner:
| Remote access to UGG company resources is restricted to AUTHORIZED USERS!
| 
| Company resources may only be used for the purpose of performing work-related duties.
| The user is responsible for protecting the confidentiality and integrity of the      company resources.
| Violation may cause disciplinary action.
| By clicking on "Continue" the user agrees with the rules above.
| 

 RTNETLINK answers: File exists
 **vpnc: quick mode response rejected:  (ISAKMP_N_INVALID_PAYLOAD_TYPE)(1)**
 this means the concentrator did not like what we had to offer.

 Possible reasons are:

 * concentrator configured to require a firewall
   this locks out even Cisco clients on any platform expect windows
   which is an obvious security improvment. There is no workaround (yet).
* concentrator configured to require IP compression
   this is not yet supported by vpnc.
   Note: the Cisco Concentrator Documentation recommends against using
   compression, expect on low-bandwith (read: ISDN) links, because it
   uses much CPU-resources on the concentrator

I also try to do the follwoing

  Add the line "NAT Traversal Mode cisco-udp" in /etc/vpnc.conf

Or

   Add the line "Local Port 10000" in /etc/vpnc.conf

But they dosent help and I still get the bad messages ( chuld not be connected by the vpnc )

Please advice where is the problem ?

When I run netstat -n on Solaris I get different results from netstat -n on Linux, even though according to the manual pages in Solaris and Linux the -n flag should do the same thing.

My question:

Please advise me how to run netstat in Linux so that I will get the same results as I get in Solaris.

EXAMPLE:

Solaris

 netstat -n | head-10

TCP:IPv4
LocalAddress       RemoteAddress   SwindSend-QRwindRecv-Q   State
 -------------------------------------------------------------------------
 10.119.99.77.9495   10.106.97.118.60823  5888     049640     0CLOSE_WAIT
 10.119.99.77.33121  10.119.99.77.1521   49152     049152     0ESTABLISHED
 10.119.99.77.1521   10.119.99.77.33121  49152     049152     0ESTABLISHED
 10.119.99.77.1521   10.119.99.112.54374 63702     049232     0ESTABLISHED
 127.0.0.1.45459     127.0.0.1.8886      49152     049152     0ESTABLISHED
 10.119.99.77.7777   10.115.244.53.63575  5888     049220     0CLOSE_WAIT

Linux

 netstat -n | head -10
 Active Internet connections (w/o servers)
 Proto Recv-Q Send-Q Local Address               Foreign Address             State      
 tcp        0      0 127.0.0.1:14407             127.0.0.1:20281             ESTABLISHED 
 tcp        0      0 10.106.207.143:61623        10.106.207.143:1521         ESTABLISHED 
 tcp        0      0 10.106.207.143:61630        10.106.207.143:1521         ESTABLISHED 
 tcp        0      0 10.106.207.143:61624        10.106.207.143:1521         ESTABLISHED 
 tcp        0      0 10.106.207.143:61625        10.106.207.143:1521         ESTABLISHED 
 tcp        0      0 10.106.207.143:61627        10.106.207.143:1521         ESTABLISHED 
 tcp        0      0 10.106.207.143:1521         10.106.207.143:32263        ESTABLISHED 
 tcp        0      0 127.0.0.1:20281             127.0.0.1:14407             ESTABLISHED

I use the following scp syntax in order to transfer allot of files from Linux red-hat 5 to windows machine ( under Temp directory ),

Remarks:

• SSH server already installed on windows machine

• I use this line in my shell scripts

   sshpass -p '$password'  /usr/bin/scp -o StrictHostKeyChecking=no  $FILE [email protected]:'D:/Temp'
  

on most cases files transferred successfully

but sometimes scp stuck during file transferring ? , in spite connectivity is ok like ping etc

and I get the following error from scp ( after a long time )

   ssh_exchange_identification: read: Connection reset by peer

1. why scp isn’t stable and stuck ? , and what the solution for this problem?
2. What the others good alternative for scp ? , (consider that I need 100% of stability)

according to ln manual page:

   man ln


   -f, --force
          remove existing destination files

so as I understand if I want to re-create new link to some destination directory/file

I can simply do

    ln -s -f some_directory new_link

but this is not what I got from my linux/solaris machines

for example

  [u@h w]# mkdir dir1
  [u@h w]# ln -s dir1 link
  [u@h w]# ls -l


  drwxr-xr-x 2 root root 4096 Mar 16 20:26 dir1
  lrwxrwxrwx 1 root root    4 Mar 16 20:27 link -> dir1




   [u@h w]# ln -s -f dir1 new_link

now I accepted to see that new_link is now pointed to dir1 while "link" isn’t exists but link still pointed to dir1 also?

how to recreate new link and on the same time remove the old link in one command ?

( I not want to use rm command )

  [u@h w]# ls -l
  drwxr-xr-x 2 root root 4096 Mar 16 20:26 dir1
  lrwxrwxrwx 1 root root    4 Mar 16 20:27 link -> dir1
  lrwxrwxrwx 1 root root    4 Mar 16 20:28 new_link -> dir1

Other different example

  # 
  # mkdir dir
  # 
  # ln -s dir link
  # 
  # mv dir dir_new
  # ls -l
  total 32
    drwxr-xr-x   2 root     root         117 Mar 16 21:44 dir_new
    lrwxrwxrwx   1 root     root           3 Mar 16 21:44 link -> dir
  # ln -s -f  dir_new link
  ln: cannot create link: File exists

is it possible to update the current link to point on new directory/file

( without to get File exists ? ) ,

or force to create the link again on different directory ?

I want to add the PATH - /usr/local/bin/perl befpore string in the file by perl command line ( the perl command line must be in my bash script ) , I have solaris machine

the target to add - /usr/local/bin/perl before the line $APIDIR/scan.pl in the file and after the char - "

please advice how to add the PATH - /usr/local/bin/perl before $APIDIR/scan.pl so finally I will get

the new line's -

      my $APIDIR="/usr/local/cp/api";           remark -  this line shouldn’t be change
      my $script="/usr/local/bin/perl   $APIDIR/scan.pl";  

in place of

     my $APIDIR="/usr/local/cp/api";
     my $script="$APIDIR/scan.pl";

I try the following but its not change the line:

  [root@machine1a /var/tmp]#  perl -p -i -e 's/\/usr\/local\/bin\/perl \$APIDIR\/scan.pl/\$APIDIR\/scan.pl/g' file




  [root@machine1a /var/tmp]# more file
  my $APIDIR="/usr/local/cp/api";
  my $script="$APIDIR/scan.pl";

I have linux machine red-hat 5.1

how to replace only the right single char (most side of string)

example from my sed syntax (not good because its replaced both "a" chars) l

 echo machine1a | sed s'/a/b/g'

 mbchine1b

But the requested answer should be - machine1b and not mbchine1b

Linux system running Redhat 5.1.

When I connect to the machines with SSH there is a long delay before the login completes and I get a shell.

I did some research and found a solution to this problem that suggested that I could remove the /etc/resolv.conf, and then logins would work quickly. I tried doing this and it actually worked; removing the /etc/resolv.conf sped things up.

So now I want to understand why this is causing a problem, and how to get fast SSH logins without breaking name resolution.

Update: UseDNS no was in sshd_config (but has not fixed the issue)

The target of the following simple expect script is to get the hostname name on the remote machine

Sometimes expect script fail to perform ssh to $IP_ADDRESS ( because remote machine not active , etc )

so in this case the expect script will break after 10 second (timeout 10) , this is OK but......

There are two options

1. Expect script perform ssh successfully , and performed the command hostname on the remote machine
2. Expect script break because timeout was 10 seconds

On both cases expect will exit

• in case of ssh successfully expect will break after 0.5-1 second but in case of bad ssh then it will break after 10 seconds

but I don’t know if expect script perform ssh successfully or not?

is it possible to identify timeout process ? or to verify that expect ended because timeout?

Remark my Linux machine version - red-hat 5.1

Expect script

 [TestLinux]# get_host_name_on_remote_machine=`cat << EOF
  > set timeout 10
  > spawn  ssh   $IP_ADDRESS
  >            expect {
  >                      ")?"   { send "yes\r"  ; exp_continue  }
  > 
  >                      word:  {send $PASS\r}
  >                   }
  > expect >  {send "hostname\r"}
  > expect >    {send exit\r}
  > expect eof
  > EOF`

Example in case we not have connection to the remote host

 [TestLinux]# expect -c  "$get_host_name_on_remote_machine"
 spawn ssh 10.17.180.23
 [TestLinux]# echo $?
 0

I run the vxprint -g rootdg on my Solaris machine (HW- NETRA T2000)

As we see here - rootvol & rootdisk7vol has missing lines (swapvol is OK)

How to fix the rootvol & rootdisk7vol in order to bring the missing lines as we see then in swapvol ?

  vxprint -g rootdg
  TY NAME         ASSOC        KSTATE   LENGTH   PLOFFS   STATE    TUTIL0  PUTIL0
  dg rootdg       rootdg       -        -        -        -        -       -

  dm rootdisk     c0t1d0s2     -        286617216 -       -        -       -
  dm rootmirror   c0t3d0s2     -        286617216 -       -        -       -

  v  rootdisk7vol gen          ENABLED  122112   -        ACTIVE   -       -
  pl rootdisk7vol-01 rootdisk7vol ENABLED 122112 -        ACTIVE   -       -
  sd rootdisk-03  rootdisk7vol-01 ENABLED 122112 0        -        -       -


  v  rootvol      root         ENABLED  253016064 -       ACTIVE   -       -
  pl rootvol-01   rootvol      ENABLED  253016064 -       ACTIVE   -       -
  sd rootdisk-02  rootvol-01   ENABLED  253016064 0       -        -       -


  v  swapvol      swap         ENABLED  33479040 -        ACTIVE   -       -
  pl swapvol-01   swapvol      ENABLED  33479040 -        ACTIVE   -       -
  sd rootdisk-01  swapvol-01   ENABLED  33479040 0        -        -       -
  pl swapvol-02   swapvol      ENABLED  33479040 -        ACTIVE   -       -
  sd rootmirror-01 swapvol-02  ENABLED  33479040 0        -        

I write the active.ksh script (based on expect) in order to login automatically to some Solaris machine and execute the hostname command (login to virtual IP in order to verify which hostname is the active machine - I have two cluster solaris machines )

The problem is with expect; expect sends the password string (pass123) and it misses the Password question, and it still waits for the password.

So actually the password (pass123) was entered after password question. On most cases the expect script works fine but sometimes it missed the password.

EXAMPLE OF THE PROBLEM

 ./active.ksh
 spawn ssh 10.10.18.61
 sh: /usr/local/bin/stty: not found
 This computer system, including all related equipment, networks and network devices      (specifically including Internet access),is provided only for authorized uss
 Password:        * my remark - pass123 string was missed the Password Question        pass123
 Password: 

THE SCRIPT

  #!/bin/ksh

  VIP_ADDRESS=10.10.18.61


  expect_for_verify_which_active_machine=`cat << EOF
  set timeout -1
  spawn  ssh   $VIP_ADDRESS 
  expect {
  ")?"   { send "yes\r"  ; exp_continue  }
  Password:  {send "pass123\r"}
  }
  expect >  {send "hostname\r"}
  expect >    {send exit\r}
  expect eof
  EOF`


  expect -c  "$expect_for_verify_which_active_machine"

EXAMPLE OF RIGHT RESULTS

  ./active.ksh 
  [Friday, February 24, 2012  2:32:06 PM IST] INFO Verify which is active SMU machine 
  spawn ssh 10.10.18.61
  sh: /usr/local/bin/stty: not found
  This computer system, including all related equipment, networks and network devices       (specifically including Internet access),is provided only for authorized uss
  yes
  Password: 
  Last login: Fri Feb 24 14:32:06 2012 from smu1a
  This computer system, including all related equipment, networks and network devices       (specifically including Internet access),is provided only for authorized uss
  solaris1:/ ROOT > hostname
  solaris1
  solaris1:/ ROOT > exit

  logout
  Connection to 10.10.18.61  closed.

I installed two Linux machines with Linux version - 5.5.1.0-01

Something is very strange if I perform ssh between Linux to Linux after I type the password I get the prompt of the second Linux machine without any delay

But if I perform ssh from window VIA (putty/reflection) to the Linux machines , after I type the password I have a long delay like 6 seconds

so please advice why I get delay when I connect to linux machine from windows

and I not get any delay when I perform ssh between Linux machines

Remark I have two Linux machine – 5.5 version + PC with windows XP all machines connected to the switch with 1G speed

I have 47 Linux client machines and one Solaris server machine (all Linux client machines are red-hat 5.1)

all clients perform ssh connection to the server in order to verify if /tmp/generic_error.txt exist in the server

clients do the ssh exactly on 12:00 PM

I see something strange

Some of clients not recognized that /tmp/generic_error.txt exist in the server in spite this file was exist

So my first conclusion is that I can't perform multiple ssh from 47 machine to one server ?

But I don’t sure if I right ?

Please advice what the best process to perform ssh from 47 Linux clients to one server exactly on the same time?

remark - I can’t touch and configure the "ssh key-based authentication" because the Lenox’s machines must have password (part of our security policy)

• example of the expect script in the linux's clients - this expect verify if /tmp/generic_error.txt exists in the server (this action should be performed exactly on 12:00 PM on all 47 linux clients)

10.10.18.61 - server machine

 expect=`cat << EOF

 set timeout -1
 spawn  ssh  10.10.18.61
       expect {
                 ")?"   { send "yes\r"  ; exp_continue  }

                 word:  {send freenelsonmandela\r}
              }
 expect >  {send "ls /tmp/generic_error.txt\r"}
 expect >    {send exit\r}
 expect eof
 EOF`

exe expect:

expect -c  "$expect"  | grep "generic_error.txt"  | grep -v ls

my target is to match Exactly the string snmpmanager from hosts file on solaris & linux

the following command work on Linux (red-hat 5.1) but not for SunOS , please advice how to fit the syntax to solaris?

example from solaris OS

   grep -icE '(^|[[:space:]])snmpmanager($|[[:space:]])' /etc/hosts
   grep: illegal option -- E

after I fixed it to

       egrep -i '(^|[[:space:]])snmpmanager($|[[:space:]])'  /etc/hosts
  or   egrep -i '(^|[\s])snmpmanager($|\s])'  /etc/hosts
  or   egrep -i '(^|[\t])snmpmanager($|\t])'  /etc/hosts

but I don’t get any match output (but snmpmanager already defined in host file) ??

my host file

     10.170.10.5      loghost
     10.170.10.61   Master SyslogSer vip Tcc NtpServer1 NtpServer2 snmpManager snmpManagerPA1 snmpManagerPA2

in my linux machine red-hat 5.1 - in ppp file - I have only the signature.pl string !!!

• remark - third application write the signature.pl string in to ppp file

I set ppp file in $a param And compare $a with signature.pl

But as all see here this not equal - why ? or if I have empty space? How to solve this?

    # more /var/tmp/ppp
    signature.pl
    # a=`cat /var/tmp/ppp`
    # echo $a
    signature.pl
    # [[ $a = signature.pl ]] && echo equal

or

    # [[ $a == signature.pl ]] && echo equal

diff example that works ! - from ksh shell (LINUX RED-HAT 5.1)

[u@h w]# echo signature.sh > file
[u@h w]# cat file
signature.sh
[u@h w]# a=`cat file`  
[u@h w]# echo $a
signature.sh
[u@h w]# [[ $a = signature.sh ]] && echo eq

 its print "eq"

I need to create watch dog process (will run in linux version 5.x) that look all time on /etc/cluster.cf file

And if the process matches the string: machineA_is_active in the cluster.cf file

Then this process will execute other script

My question is – how to run this process in way that process will up and running all time that Linux is up –

and in case this process is down need to startup this process again

So please advice what the basic structure for this scenario ?

(I will happy if I get real example)

please advice how to solve this problem I have linux with informix DB its fail on - Fatal error in shared memory creation

su - informix

-sh-2.05b$ oninit -v

    ONCONFIG /usr/informix/etc/onconfig.active is not owned by user with id 4000.
    Checking group membership to determine server run modesucceeded
    Reading configuration file '/usr/informix/etc/onconfig.active'...succeeded
    Creating /INFORMIXTMP/.infxdirs ... succeeded
     Creating infos     file "/usr/informix/etc/.infos.active_trm" ... "/usr/informix/etc/.conf.active_trm" ... s succeeded
    Writing to infos file ... succeeded
    Checking config parameters...succeeded
    Allocating and attaching to shared memory...FAILED
    oninit: Fatal error in shared memory creation

oninit -i

  This action will initialize IBM Informix Dynamic Server;
   any existing IBM Informix Dynamic Server databases will NOT be accessible -
   Do you wish to continue (y/n)? 
   please enter 'y' or 'n'...
   Do you wish to continue (y/n)? y
   oninit: Fatal error in shared memory creation

Usually if we want to redirect script output to file we simply do

   /etc/create_connection.sh > file.debug.txt  

Or

   /etc/create_connection.sh | tee –a  file.debug.txt  

But What I want is to add /etc/create_connection.sh > file.debug.txt inside the script , but I not sure if it possible and how?

so when I run just

  /etc/create_connection.sh 

I need that all standard output will write to file.debug.txt
so what I need to add in my script in order to do that?

I run the following sed syntax on Solaris machine and on Linux machine

on Linux machine sed do the JOB its remove all characters until the first number.

but why this sed syntax not work on Solaris ?

What I need to change in the sed syntax in order run it on Solaris?

on solaris (sed - not remove the strings until the first number?)

   solaris:/ ROOT > echo "Release............5.3.7.1-12"  | sed 's/[^0-9]\+//'

   Release............5.3.7.1-12

on linux ( I get good results )

  linux tmp]# echo "Linux Release............5.3.7.1-12"  | sed 's/[^0-9]\+//'

  5.3.7.1-12

how to enable find command to print exe status diff from 0 when find command not find the file?

according to my example when I try to find the test1.txt file , $? set to 0 but I expect to get value diff from 0 because the file not found

 [root@om-1 tmp]# touch test.txt
 [root@om-1 tmp]# find /var/tmp -name  test.txt
 /var/tmp/test.txt
 [root@om-1 tmp]# echo $?
                  0
 [root@om-1 tmp]# find /var/tmp -name  test1.txt (test1.txt not under /var/tmp)
 [root@om-1 tmp]# echo $?
                  0

I have the following line in a file:

      Linux Release............5.4.2.0-02 12_12_2011_07:31:23

How do I remove all characters before the first number with sed or awk?

I wish to get the following result:

      5.4.2.0-02 12_12_2011_07:31:23