Fence na: Difference between revisions

From Alteeve Wiki
Jump to navigation Jump to search
No edit summary
No edit summary
 
(5 intermediate revisions by the same user not shown)
Line 3: Line 3:
This is the core fence agent that exists in <span class="code">/sbin/</span>.
This is the core fence agent that exists in <span class="code">/sbin/</span>.


* Download the [http://nodeassassin.org/files/sbin/fence_na source code].
<source lang="perl">
<source lang="perl">
#!/usr/bin/perl
#!/usr/bin/perl
Line 8: Line 9:
# Node Assassin - Fence Agent
# Node Assassin - Fence Agent
# Digimer; digimer@alteeve.com
# Digimer; digimer@alteeve.com
# Mar. 07, 2010.
# Jun. 27, 2010
# Version: 0.1.004
# Version: 1.1.5
#
# This software is released under the GPL v2. See the LICENSE file in the
# configuration directory for a copy of the GPL v2.
#
#
# Bugs;
# Bugs;
# - None known, many expected
# - None known, many expected
#  
#  
=pod
Changes:
v0.1.004
- Fixed the command line argument bug.
- Updated the 'help' message to be more accurate.
Given the following:
<cluster name="an_san" config_version="1">
<clusternodes>
<clusternode name="an_san01.alteeve.com" nodeid="1">
<fence>
<method name="node_assassin">
<device name="ariel" port="01" action="off"/>
</method>
</fence>
</clusternode>
<clusternode name="an_san02.alteeve.com" nodeid="2">
<fence>
<method name="node_assassin">
<device name="ariel" port="02" action="off"/>
</method>
</fence>
</clusternode>
</clusternodes>
<fencedevices>
<fencedevice name="node_assassin" agent="fence_na" ipaddr="ariel.alteeve.com" name="ariel" passwd="gr0tt0"></fencedevice>
</fencedevices>
</cluster>
Questions:
- Is there a corelation between 'clusternode -> name', 'device -> name' and
'fencedevice -> name'? Which is used when sending 'name' to the fence agent?
'fencedevice'?
 
When 'fenced' decides to fence "an_san01.alteeve.com", it will:
- call '/sbin/fence_na' because of the 'fencedevices -> agent' value.
- It will pass the following arguments to the fence agent, one pair per line:
    agent=fence_na # From 'fencedevices -> agent'
    name=ariel # From 'fencedevices -> name'
    ipaddr=ariel.alteeve.com # From 'fencedevices -> ipaddr'
    passwd=gr0tt0 # From 'fencedevices -> passwd'
    port=01 # From 'clusternode "an_san01.alteeve.com" -> port'
    action=fence_na # From 'clusternode "an_san01.alteeve.com" -> option'
    # This must be 'on', 'off', 'reboot', 'status'
    # or 'monitor'. See below for how these terms
    # are interpretted by this agent.
    # NOTE: If 'option' is passed, it's value will
    # be stored in 'action'. That is, 'action' and
    # 'option' are synonymous.
    #
- Node Assassin's implentation of options.
  - 'off' This sets the node to state '0' on the reset port followed by
  state '3' to the power port. State 0 is maintained to prevent
  a reboot.
  - 'on' This sets the node to state '1' on the reset port followed by
  state '2' on the power port to boot the node.
  - 'reboot' This sets the node to state '2' on the reset port to quickly
  kill the node, then switches to state '3' on the power port,
  checks the return value (later, will check the probe pin),
  sets state '1' on the reset port, pauses 1 second, and then
  sets state '2' on the power port to boot the node.
  - 'status' This calls '00:0' and returns the state of the port. Later,
  this will return the value from the voltage sensing pin.
  - 'monitor' being a multi-port fence device, this should call 'list'.
  MADI: Confirm that this is what is meant in "Issues" here:
  http://sources.redhat.com/cluster/wiki/FenceAgentAPI
  - 'list' No info on this
Command Line Arguments:
- Any command line arguments used by this fence agent are not dictated by the
  Fence Agent API. By convention only, the following command line options are
  used:
  -a <ip> # Maps the value to 'ipaddr'.
  -h # Print the help message and then exits.
  -l <name> # Maps the value to 'name'.
  -n <num> # Maps the value to 'port'.
  -o <string> # Maps the value to 'action'.
  -p <string> # Maps the value to 'passwd'.
  -S <path> # Maps the value to 'passwd_script'. This is not used by Node
  # Assassin yet and is simply ignored.
  -q # Sets quiet mode. Only errors will be printed. Logging
  # proceeds as normal
  -V # Prints the 'fence_na' version and the version of any attached
  # Node Assassin(s) and exits.
Note:
- For now, I will return '0' if the command succeeded, but will add a detection
  line checks if there is voltage from the node's PSU later.
=cut


# Play safe!
# Play safe!
use strict;
use strict;
use warnings;
use warnings;
# Load my library.
 
require '/etc/na/fence_na.lib';
# Load our library.
# This is how I talk.
require '/etc/fence_na/fence_na.lib';
 
# IO::Handle is used for logging and Net::Telnet is used for communicating with
# the Node Assassin(s).
use IO::Handle;
use IO::Handle;
use Net::Telnet;
use Net::Telnet;


# This will be read in from a config file later.
# These are the default values and will be over-written by the config file's
# variables which in turn can, in some cases, be over-written by command line
# arguments.
# Please see '/etc/na/fence_na.conf' for details on each option.
my $conf={
my $conf={
'system' => {
'system' => {
max_valid_state => 3,
max_valid_state => 3,
conf_file => "/etc/na/fence_na.conf",
conf_file => "/etc/fence_na/fence_na.conf",
quiet => "",
quiet => "",
version => "",
version => 0,
list_state => "",
list => "",
list => "",
monitor => "",
monitor => "",
node_assassin_id=> 0,
na_id => 0,
got_cla => 0, # This is set if command line arguments are read.
got_cla => 0, # This is set if command line arguments are read.
debug => 0,
},
},
node => {
na => {
ipaddr => "",
ipaddr => "",
tcp_port => "",
tcp_port => "238",
port => "238",
port => "",
login => "",
login => "",
passwd => "",
passwd => "",
Line 142: Line 58:
action => "",
action => "",
agent => "", # This is only used by 'fenced'
agent => "", # This is only used by 'fenced'
na_name => "", # This is used for the 'list' function.
handle => "",
handle => "",
set_state => [], # This anon array will store the states to set based on the action passed for the proper ports.
max_node => 0,
set_state => [], # This array will store the states to set based on the action passed for the proper ports.
}
}
};
};
# This method can't pass in the '$log' handle, obviously, as it does not yet
# This method can't pass in the '$log' handle, obviously, as it does not yet
# exist.
# exist.
Line 152: Line 71:
# Log file for output.
# Log file for output.
my $log=IO::Handle->new();
my $log=IO::Handle->new();
print "Opening: [$conf->{'system'}{'log'}] for logging.\n"  if $conf->{'system'}{debug};
open ($log, ">$conf->{'system'}{'log'}") || die "Failed to open: [$conf->{'system'}{'log'}] for writing; Error: $!\n";
open ($log, ">$conf->{'system'}{'log'}") || die "Failed to open: [$conf->{'system'}{'log'}] for writing; Error: $!\n";
# Set STDOUT and $log to hot (unbuffered) output.
if (1)
{
select $log;
$|=1;
select STDOUT;
$|=1;
}


# If this gets set in the next two function, the agent will exit.
# If this gets set in the next two function, the agent will exit.
Line 163: Line 92:
($bad)=read_stdin($conf, $log, $bad);
($bad)=read_stdin($conf, $log, $bad);


# This makes sure the node ID is zero-padded or '00'.
# This makes sure the node ID is either zero-padded or '00'.
$conf->{node}{port}=$conf->{node}{port} ? $conf->{node}{port}=sprintf("%02d", $conf->{node}{port}) : "00";
$conf->{na}{port}=$conf->{na}{port} ? $conf->{na}{port}=sprintf("%02d", $conf->{na}{port}) : "00";
record($conf, $log, "Will use port: [$conf->{na}{port}]\n") if $conf->{'system'}{debug};


# Find the TCP port from the config file.
# Find the TCP port from the config file.
foreach my $i (1..$conf->{'system'}{na_num})
foreach my $i (1..$conf->{'system'}{na_num})
{
{
if ((lc($conf->{node}{$i}{ipaddr}) eq lc($conf->{node}{ipaddr})))
if ((lc($conf->{na}{$i}{ipaddr}) eq lc($conf->{na}{ipaddr})))
{
{
$conf->{'system'}{node_assassin_id}=$i;
$conf->{'system'}{na_id}=$i;
$conf->{node}{tcp_port}=$conf->{node}{$i}{tcp_port};
record($conf, $log, __LINE__."; system::na_id: [$conf->{'system'}{na_id}]\n") if $conf->{'system'}{debug};
last;
$conf->{na}{tcp_port}=$conf->{na}{$i}{tcp_port};
record($conf, $log, __LINE__."; na::tcp_port: [$conf->{na}{tcp_port}]\n") if $conf->{'system'}{debug};
$conf->{na}{na_name}=$conf->{na}{$i}{na_name} ? $conf->{na}{$i}{na_name} : "Node Assassin #$i";
record($conf, $log, __LINE__."; na::na_name: [$conf->{na}{na_name}]\n") if $conf->{'system'}{debug};
$conf->{na}{max_nodes}=$conf->{na}{$i}{max_nodes};
record($conf, $log, __LINE__."; na::max_nodes: [$conf->{na}{max_nodes}]\n") if $conf->{'system'}{debug};
}
}
}
}


die "Exiting on errors.\n" if $bad;
die "Exiting on errors.\n" if $bad;
record($conf, $log, "Node Assassin: [$conf->{node}{ipaddr}].\n");
my @ny=("no", "yes");
record($conf, $log, "TCP Port: .... [$conf->{node}{tcp_port}].\n");
record($conf, $log, "Node Assassin: . [$conf->{na}{ipaddr}].\n");
record($conf, $log, "Port: ........ [$conf->{node}{port}].\n");
record($conf, $log, "TCP Port: ...... [$conf->{na}{tcp_port}].\n");
record($conf, $log, "Login: ....... [$conf->{node}{login}].\n");
record($conf, $log, "Node: .......... [$conf->{na}{port}].\n");
record($conf, $log, "Password: .... [$conf->{node}{passwd}].\n");
record($conf, $log, "Login: ......... [$conf->{na}{login}].\n");
record($conf, $log, "Action: ...... [$conf->{node}{action}].\n");
record($conf, $log, "Password: ...... [$conf->{na}{passwd}].\n");
record($conf, $log, "Action: ........ [$conf->{na}{action}].\n");
record($conf, $log, "Version Request: [".$ny[$conf->{'system'}{version}]."].\n");
record($conf, $log, "Done reading args.\n");
record($conf, $log, "Done reading args.\n");
# If I've been asked to show the version information, do so and then exit.
record($conf, $log, "Version: ..... [$conf->{'system'}{version}].\n") if $conf->{'system'}{debug};
if ($conf->{'system'}{version})
{
version($conf, $log);
do_exit($conf, $log, 0);
}


# Connect to the Node Assassin.
# Connect to the Node Assassin.
$conf->{node}{handle}=new Net::Telnet(
connect_to_na($conf, $log);
Timeout => 10,
Errmode => 'die',
Port => $conf->{node}{tcp_port},
Prompt => '/EOM$/',
Errmode => 'return'
) or do_exit($conf, $log, 1);
# print "Handle: [$conf->{node}{handle}]\n";
$conf->{node}{handle}->open($conf->{node}{ipaddr});


# Validate credentials.
# Validate credentials.
# NOTE: Checking before the telnet fails on the exit. Also, this will be moved
# NOTE: Checking before the telnet fails on the exit. Also, this will be moved
# into the Node Assassin soon anyway.
# into the Node Assassin soon anyway.
if (($conf->{node}{login} ne $conf->{'system'}{username}) or ($conf->{node}{passwd} ne $conf->{'system'}{password}))
if (($conf->{na}{login} ne $conf->{'system'}{username}) or ($conf->{na}{passwd} ne $conf->{'system'}{password}))
{
{
record($conf, $log, "Username and/or password failed.\n");
record($conf, $log, "Username and/or password invalid. Did you use the command line switches properly?\n");
do_exit($conf, $log, 8);
do_exit($conf, $log, 8);
}
}
Line 210: Line 147:
###############################################################################
###############################################################################


# If I've been asked to show the version information, do so and then exit.
# When asked to 'monitor' or 'list'. being multi-port, this will return a CSV
record($conf, $log, "Version: ..... [$conf->{'system'}{version}].\n");
# of nodes and their aliases where found in the config file.
if ($conf->{'system'}{version})
record($conf, $log, "Action: ........ [$conf->{na}{action}].\n") if $conf->{'system'}{debug};
{
if (($conf->{na}{action} eq "monitor") or ($conf->{na}{action} eq "list"))
version($conf, $log);
do_exit($conf, $log, 0);
}
 
# If I've been asked to show the info on the given node assassin, do so and
# then exit.
record($conf, $log, "List State: .. [$conf->{'system'}{list_state}].\n");
if ($conf->{'system'}{list_state})
{
show_state($conf, $log);
do_exit($conf, $log, 0);
}
 
# When asked to 'monitor' or 'list', do this... whatever 'this' is. All I know
# is that it should not generate output.
record($conf, $log, "Monitor: ..... [$conf->{'system'}{monitor}].\n");
record($conf, $log, "List: ........ [$conf->{'system'}{list}].\n");
if (($conf->{node}{monitor}) or ($conf->{node}{list}))
{
{
record($conf, $log, "Calling the 'show_list' function.\n") if $conf->{'system'}{debug};
show_list($conf, $log);
show_list($conf, $log);
do_exit($conf, $log, 0);
do_exit($conf, $log, 0);
Line 238: Line 158:


# If I made it this far, I am setting a state. Sort out what state from the
# If I made it this far, I am setting a state. Sort out what state from the
# values in my conf->{node} hash.
# values in my conf->{na} hash.
record($conf, $log, "Setting node: [$conf->{node}{port}] to action: [$conf->{node}{action}] using the Node Assassin: [$conf->{node}{ipaddr}] using the login: [$conf->{node}{login}/$conf->{node}{passwd}]\n");
record($conf, $log, "Setting node: [$conf->{na}{port}] to action: [$conf->{na}{action}] using the Node Assassin: [$conf->{na}{ipaddr}] using the login: [$conf->{na}{login}]\n") if $conf->{'system'}{debug};


# Convert the action into Node Assassin protocol arguments.
# Convert the action into Node Assassin protocol arguments.
process_action($conf, $log);
process_action($conf, $log);


# In the next step, when a 'check' is seen, the port is analyzed and an exit
# Now execute the action plan.
# status is stored here. Exits 0, 1 and 2 have special meaning, so I default to
my $exit_code=do_actions($conf, $log);
# 9.
record($conf, $log, "All calls complete, exiting.\n") if $conf->{'system'}{debug};
my $exit_code=9;
 
# Process the orders.
foreach my $order (split/,/, $conf->{'system'}{call_order})
{
if ($order=~/^sleep/)
{
my $time=$order=~/sleep (\d+)/ ? $1 : 1;
record ($conf, $log, "Sleeping for: [$time]...\n");
sleep $time;
next;
}
record ($conf, $log, "Calling: [$order]\n");
if ($order=~/(\d\d):(\D+)/)
{
my $node=$1;
my $check=$2;
# Verify the state of the port.
record($conf, $log, "Status check on node: [$node] -> [$check]\n");
# Get the state.
my $states=get_states($conf, $log);
if ($states == 1)
{
# I had a connection problem.
do_exit($conf, $log, 1);
}
my $node_state=$states->{$node};
record($conf, $log, "Node: [$node] state: [$node_state]\n");
if ($check eq "check")
{
# Return '2' if the node is off and '0' if it is on.
$exit_code=$node_state eq "off" ? 2 : 0;
}
elsif ($check eq "off")
{
# 'off' was called, make sure the node is now off. This
# may be called by 'reboot' in which case 'exit_code'
# will simply be over-written when the final 'reboot'
# state check is called.
$exit_code=$node_state eq "off" ? 0 : 1;
}
elsif ($check eq "on")
{
# 'on' was called, make sure the node is now off.
$exit_code=$node_state eq "off" ? 1 : 0;
}
elsif ($check eq "reboot")
{
# Make sure that 'exit_code' was set to '0' by the
# earlier call. We checked again to make sure the node
# came back up, and will log an error if it didn't, but
# we return '0' just the same, as per the API.
if ($exit_code eq "0")
{
# The power off portion worked. Check if the
# node booted properly and record an error if
# not.
if ($node_state eq "off")
{
record($conf, $log, "Node: [$node] failed to boot after a successful power off during a reboot action.\n");
record($conf, $log, "This is a non-critical error as the node was fenced successfully but may\n");
record($conf, $log, "indicate a hardware failure with the node or with Node Assassin itself.\n");
}
}
else
{
# The power off portion failed, exit with '1'.
$exit_code=1;
}
$exit_code=$node_state eq "off" ? 1 : 0;
}
next;
}
my @set_state=$conf->{node}{handle}->cmd("$order");
foreach my $line (@set_state)
{
record($conf, $log, $line);
}
record($conf, $log, "Call complete.\n");
}
record($conf, $log, "All calls complete, exiting.\n");
 
# Now confirm that the requested node is in it's requested state and exit with
# the appropriate exit code. This function should not return.
# show_state($conf, $log);


# Cleanup and exit.
# Cleanup and exit.

Latest revision as of 17:03, 25 August 2010

 Node Assassin :: Fence na

This is the core fence agent that exists in /sbin/.

#!/usr/bin/perl
#
# Node Assassin - Fence Agent
# Digimer; digimer@alteeve.com
# Jun. 27, 2010
# Version: 1.1.5
#
# This software is released under the GPL v2. See the LICENSE file in the
# configuration directory for a copy of the GPL v2.
#
# Bugs;
# - None known, many expected
# 

# Play safe!
use strict;
use warnings;

# Load our library.
require '/etc/fence_na/fence_na.lib';

# IO::Handle is used for logging and Net::Telnet is used for communicating with
# the Node Assassin(s).
use IO::Handle;
use Net::Telnet;

# These are the default values and will be over-written by the config file's
# variables which in turn can, in some cases, be over-written by command line
# arguments.
# Please see '/etc/na/fence_na.conf' for details on each option.
my $conf={
	'system'	=>	{
		max_valid_state	=>	3,
		conf_file	=>	"/etc/fence_na/fence_na.conf",
		quiet		=>	"",
		version		=>	0,
		list		=>	"",
		monitor		=>	"",
		na_id		=>	0,
		got_cla		=>	0,	# This is set if command line arguments are read.
		debug		=>	0,
	},
	na		=>	{
		ipaddr		=>	"",
		tcp_port	=>	"238",
		port		=>	"",
		login		=>	"",
		passwd		=>	"",
		port		=>	"",
		set_state	=>	"",
		passwd_script	=>	"",
		action		=>	"",
		agent		=>	"",	# This is only used by 'fenced'
		na_name		=>	"",	# This is used for the 'list' function.
		handle		=>	"",
		max_node	=>	0,
		set_state	=>	[],	# This array will store the states to set based on the action passed for the proper ports.
	}
};

# This method can't pass in the '$log' handle, obviously, as it does not yet
# exist.
read_conf($conf);

# Log file for output.
my $log=IO::Handle->new();
print "Opening: [$conf->{'system'}{'log'}] for logging.\n"  if $conf->{'system'}{debug};
open ($log, ">$conf->{'system'}{'log'}") || die "Failed to open: [$conf->{'system'}{'log'}] for writing; Error: $!\n";

# Set STDOUT and $log to hot (unbuffered) output.
if (1)
{
	select $log;
	$|=1;
	select STDOUT;
	$|=1;
}

# If this gets set in the next two function, the agent will exit.
my $bad=0;

# Read in arguments from the command line.
($bad)=read_cla($conf, $log, $bad);

# Now read in arguments from STDIN, which is how 'fenced' passes arguments.
($bad)=read_stdin($conf, $log, $bad);

# This makes sure the node ID is either zero-padded or '00'.
$conf->{na}{port}=$conf->{na}{port} ? $conf->{na}{port}=sprintf("%02d", $conf->{na}{port}) : "00";
record($conf, $log, "Will use port: [$conf->{na}{port}]\n") if $conf->{'system'}{debug};

# Find the TCP port from the config file.
foreach my $i (1..$conf->{'system'}{na_num})
{
	if ((lc($conf->{na}{$i}{ipaddr}) eq lc($conf->{na}{ipaddr})))
	{
		$conf->{'system'}{na_id}=$i;
		record($conf, $log, __LINE__."; system::na_id: [$conf->{'system'}{na_id}]\n") if $conf->{'system'}{debug};
		$conf->{na}{tcp_port}=$conf->{na}{$i}{tcp_port};
		record($conf, $log, __LINE__."; na::tcp_port: [$conf->{na}{tcp_port}]\n") if $conf->{'system'}{debug};
		$conf->{na}{na_name}=$conf->{na}{$i}{na_name} ? $conf->{na}{$i}{na_name} : "Node Assassin #$i";
		record($conf, $log, __LINE__."; na::na_name: [$conf->{na}{na_name}]\n") if $conf->{'system'}{debug};
		$conf->{na}{max_nodes}=$conf->{na}{$i}{max_nodes};
		record($conf, $log, __LINE__."; na::max_nodes: [$conf->{na}{max_nodes}]\n") if $conf->{'system'}{debug};
	}
}

die "Exiting on errors.\n" if $bad;
my @ny=("no", "yes");
record($conf, $log, "Node Assassin: . [$conf->{na}{ipaddr}].\n");
record($conf, $log, "TCP Port: ...... [$conf->{na}{tcp_port}].\n");
record($conf, $log, "Node: .......... [$conf->{na}{port}].\n");
record($conf, $log, "Login: ......... [$conf->{na}{login}].\n");
record($conf, $log, "Password: ...... [$conf->{na}{passwd}].\n");
record($conf, $log, "Action: ........ [$conf->{na}{action}].\n");
record($conf, $log, "Version Request: [".$ny[$conf->{'system'}{version}]."].\n");
record($conf, $log, "Done reading args.\n");

# If I've been asked to show the version information, do so and then exit.
record($conf, $log, "Version: ..... [$conf->{'system'}{version}].\n") if $conf->{'system'}{debug};
if ($conf->{'system'}{version})
{
	version($conf, $log);
	do_exit($conf, $log, 0);
}

# Connect to the Node Assassin.
connect_to_na($conf, $log);

# Validate credentials.
# NOTE: Checking before the telnet fails on the exit. Also, this will be moved
# into the Node Assassin soon anyway.
if (($conf->{na}{login} ne $conf->{'system'}{username}) or ($conf->{na}{passwd} ne $conf->{'system'}{password}))
{
	record($conf, $log, "Username and/or password invalid. Did you use the command line switches properly?\n");
	do_exit($conf, $log, 8);
}

###############################################################################
# What do?                                                                    #
###############################################################################

# When asked to 'monitor' or 'list'. being multi-port, this will return a CSV
# of nodes and their aliases where found in the config file.
record($conf, $log, "Action: ........ [$conf->{na}{action}].\n") if $conf->{'system'}{debug};
if (($conf->{na}{action} eq "monitor") or ($conf->{na}{action} eq "list"))
{
	record($conf, $log, "Calling the 'show_list' function.\n") if $conf->{'system'}{debug};
	show_list($conf, $log);
	do_exit($conf, $log, 0);
}

# If I made it this far, I am setting a state. Sort out what state from the
# values in my conf->{na} hash.
record($conf, $log, "Setting node: [$conf->{na}{port}] to action: [$conf->{na}{action}] using the Node Assassin: [$conf->{na}{ipaddr}] using the login: [$conf->{na}{login}]\n") if $conf->{'system'}{debug};

# Convert the action into Node Assassin protocol arguments.
process_action($conf, $log);

# Now execute the action plan.
my $exit_code=do_actions($conf, $log);
record($conf, $log, "All calls complete, exiting.\n") if $conf->{'system'}{debug};

# Cleanup and exit.
do_exit($conf, $log, $exit_code);

 

Input, advice, complaints and meanderings all welcome!
Digimer digimer@alteeve.ca https://alteeve.ca/w legal stuff:  
All info is provided "As-Is". Do not use anything here unless you are willing and able to take resposibility for your own actions. © 1997-2013
Naming credits go to Christopher Olah!
In memory of Kettle, Tonia, Josh, Leah and Harvey. In special memory of Hannah, Jack and Riley.