Expect script help needed- skipping unavailable routers?
Hey all. Sometimes I'm tasked to change some router configs for the entire network (over 3,000 Cisco routers). Most of the time its a global config parameter so its done with a loop and an IP list as its the same configuration change for all routers. This is working OK.
However, sometimes an expect script must be generated for ALL routers since the config changes are different for each router. The way we approach it, we divide the routers per regional domains (6 in total), and produce 1 expect script per regional domain. The problem I'm having is that when I schedule these type of expect scripts to run via crontab, they usually get hung anytime a router fails to respond (maintenance, outage, etc). How can I make the script go to the next router if a router cannot be reached? Here's a sample line of the 20K lines:
Well, I am not even sure if its failing, cause at the other end I have a select call and it wakes up and reads the data I sent fine.
Ok here is the issue, I have a UDP socket(non blocking) through which I push some data to another port. At the other end I have select loop, waiting for this data.... (6 Replies)
Hello to all...this is my first post (so please go easy). :)
I feel pretty solid at expect scripting, but I'm running into an issue that I'm not able to wrap my head around. I wrote a script that is a little advanced for logging into a remote Linux machine and changing text in a file using sed.... (2 Replies)
I'm calling an expect script via a ksh script in cron and it is failing. The script runs fine if i run it manually. Does anyone know if it is an issue with compatibilty and if there is a way around it? (2 Replies)
This Expect script provides expect with a list of IP addresses to Cisco IPS sensors and commands to configure Cisco IPS sensors. The user, password, IP addresses, prompt regex, etc. have been anonymized. In general this script will log into the sensors and send commands successfully but there are... (1 Reply)
Solaris 10 Server refuse to connect :wall:
fork: Resource temporarily unavailable , server unexpectedly unavailable network connection , refuse error, disconnect message, fatal error type2, (protocol error type2)
Issue has been resolved after taken few steps :b:
First of all need to check... (1 Reply)
Hi,
Am very new to expect scripting..
Can You please suggest me how to call an expect script inside another expect script..
I tried with
spawn /usr/bin/ksh
send "expect main.exp\r"
expect $root_prompt
and
spawn /usr/bin/ksh
send "main.exp\r"
expect $root_prompt
Both... (1 Reply)
I have an expect script called remote that I want to call from inside my expect script called sudoers.push, here is the code that is causing me issues:
set REMOTE "/root/scripts/remote"
...
log_user 1
send_user "Executing remote script as $user...\n"
send_user "Command to execute is: $REMOTE... (1 Reply)
Dear All,
I have a requirement where I have to SFTP or SCP a file in a batch script. Unfortunately, the destination server setup is such that it doesn't allow for shell command line login. So, I am not able to set up SSH keys. My source server is having issues with Expect. So, unable to use... (5 Replies)
Hy guys,
My name is Alex, i am new here and I hope to find some answers. I am trying to run a expect script to telnet to a mikrotik router, run a command (export), and save the output of that commant to a file (outputfile.txt). The problem is that only part of the output is saved to... (2 Replies)
I'm fairly new to scripting so this might not be possible.
I am using Expect with Cisco switches and need to capture the string after finding the expect request. For example, when I issue "show version" on a Nexus switch, I'm looking to capture the current firmware version:
#show version
... (0 Replies)
Discussion started by: IBGaryA
0 Replies
LEARN ABOUT CENTOS
runscript
RUNSCRIPT(1) General Commands Manual RUNSCRIPT(1)NAME
runscript - script interpreter for minicom
SYNOPSIS
runscript scriptname [logfile [homedir]]
DESCRIPTION
runscript is a simple script interpreter that can be called from within the minicom communications program to automate tasks like logging
in to a Unix system or your favorite BBS.
INVOCATION
The program expects a script name and optionally a filename and the user's home directory as arguments, and it expects that it's input and
output are connected to the "remote end", the system you are connecting to. All messages from runscript meant for the local screen are
directed to the stderr output. All this is automatically taken care of if you run it from minicom. The logfile and home directory parame-
ters are only used to tell the log command the name of the logfile and where to write it. If the homedir is omitted, runscript uses the
directory found in the $HOME environment variable. If also the logfile name is omitted, the log commands are ignored.
KEYWORDS
Runscript recognizes the following commands:
expect send goto gosub return !
exit print set inc dec if
timeout verbose sleep break call log
OVERVIEW OF KEYWORDS
send <string>
<string> is sent to the modem. It is followed by a '
'. <string> can be:
- regular text, e.g. 'send hello'
- text enclosed in quotes, e.g. 'send "hello world"'
Within <string> the following sequences are recognized:
- newline
- carriage return
a - bell
- backspace
c - don't send the default '
'.
f - formfeed
^ - the ^ character
o - send character o (o is an octal number)
Control characters can be used in the string with the ^ prefix (^A to ^Z, ^[, ^ ^], ^^ and ^_). If you need to send the ^ character,
you must prefix it with the escape character.
Also $(environment_variable) can be used, for example $(TERM). Minicom passes three special environment variables: $(LOGIN), which is
the username, $(PASS), which is the password, as defined in the proper entry of the dialing directory, and $(TERMLIN) which is the
number of actual terminal lines on your screen (that is, the statusline excluded).
print <string>
Prints <string> to the local screen. Default followed by '
'. See the description of 'send' above.
label:
Declares a label (with the name 'label') to use with goto or gosub.
goto <label>
Jump to another place in the program.
gosub <label>
Jumps to another place in the program. When the statement 'return' is encountered, control returns to the statement after the gosub.
Gosub's can be nested.
return
Return from a gosub.
! <command>
Runs a shell for you in which 'command' is executed. On return, the variable '$?' is set to the exit status of this command, so you
can subsequently test it using 'if'.
exit [value]
Exit from "runscript" with an optional exit status. (default 1)
set <variable> <value>
Sets the value of <variable> (which is a single letter a-z) to the value <value>. If <variable> does not exist, it will be created.
<value> can be a integer value or another variable.
inc <variable>
Increments the value of <variable> by one.
dec <variable>
Decrements the value of <variable> by one.
if <value> <operator> <value> <statement>
Conditional execution of <statement>. <operator> can be <, >, != or =. Eg, 'if a > 3 goto exitlabel'.
timeout <value>
Sets the global timeout. By default, 'runscript' will exit after 120 seconds. This can be changed with this command. Warning: this
command acts differently within an 'expect' statement, but more about that later.
verbose <on|off>
By default, this is 'on'. That means that anything that is being read from the modem by 'runscript', gets echoed to the screen. This
is so that you can see what 'runscript' is doing.
sleep <value>
Suspend execution for <value> seconds.
expect
expect {
pattern [statement]
pattern [statement]
[timeout <value> [statement] ]
....
}
The most important command of all. Expect keeps reading from the input until it reads a pattern that matches one of the specified
ones. If expect encounters an optional statement after that pattern, it will execute it. Otherwise the default is to just break out
of the expect. 'pattern' is a string, just as in 'send' (see above). Normally, expect will timeout in 60 seconds and just exit, but
this can be changed with the timeout command.
break
Break out of an 'expect' statement. This is normally only useful as argument to 'timeout' within an expect, because the default action
of timeout is to exit immediately.
call <scriptname>
Transfers control to another scriptfile. When that scriptfile finishes without errors, the original script will continue.
log <text>
Write text to the logfile.
NOTES
If you want to make your script to exit minicom (for example when you use minicom to dial up your ISP, and then start a PPP or SLIP session
from a script), try the command "! killall -9 minicom" as the last script command. The -9 option should prevent minicom from hanging up the
line and resetting the modem before exiting.
Well, I don't think this is enough information to make you an experienced 'programmer' in 'runscript', but together with the examples it
shouldn't be too hard to write some useful script files. Things will be easier if you have experience with BASIC. The minicom source code
comes together with two example scripts, scriptdemo and unixlogin. Especially the last one is a good base to build on for your own scripts.
SEE ALSO minicom(1)BUGS
Runscript should be built in to minicom.
AUTHOR
Miquel van Smoorenburg, <miquels@drinkel.ow.org> Jukka Lahtinen, <walker@netsonic.fi>
User's Manual $Date: 2007-10-07 18:13:51 $ RUNSCRIPT(1)