Removing docker network in swarm mode gives error below:

Error response from daemon: network demo_net has active endpoints

In order to remove the network you need to remove dangling containers that are bound to the network, to see those containers:

docker network inspect demo_net

In Containers section you would see name of those containers. Grab the name of the container and use disconnect as below to remove it from the network:

docker network disconnect -f demo_net NAME_OF_YOUR_CONTAINER

#docker #network #disconnect #swarm

How to create a network scanner in python using socket?

socket library is used to work with network and here we use it to simulate a network scanner.

A network scanner iterates over all the possible IP addresses on the network and for every IP it tries to connect to one of the ports. If the connection was successful (ACK received) we know that this host is available at the given IP address. This is the philosophy of a network scanner.

For now we just try to connect to a specific port of the server and leave the other parts to you to iterate and check the status:

import socket

  addr = '172.16.132.20'
  port = 27017

  socket_obj = socket.socket(socket.AF_INET,socket.SOCK_STREAM)
  socket.setdefaulttimeout(1)
  result = socket_obj.connect_ex((addr,port))
  print result
  socket_obj.close()

The above code connects to the given address:port and returns a status code. If returned code is 0 then the
port is open otherwise the port is probably blocked/closed.

#python #socket #network #tcp #tcp_scanner

How to monitor network cards on Icinga2? (part-1)

Yesterday I've been on a task of monitoring network cards of all our servers and infrastructure to check the bandwidth in/out and send alarms based on some criteria. In Icigna2 we have a library from nagios called check_nwc_health. Download the script from https:// labs.consol.de/nagios/check_nwc_health/index.html.

Move the script to /usr/lib/nagios/plugins on a server that you have installed Icinga2. If you run it all alone you will get some helps that you could be useful.

Some important usages of the script:

- list interfaces of a specific server (we assume snmp has been installed on the destination server):

./check_nwc_health --mode list-interfaces --hostname YOUR_TARGET_SERVER_IP  --community YOUR_COMMUNITY_STRING

The output would be something like below (it can be different in your case):

000001 lo
 000002 Device 1af4:0001 2
 000003 Device 1af4:0001 3
 000004 docker0
 OK - have fun

The interface name is given in front of serial numbers which is lo, Device 1af4:0001 2 or docker0. These interface names are important and will be used in icinga2 to add network card to hosts.


Another mode for the script is interface-usage that shows in/out bandwidth. The output can be something like follow:

OK - interface Device 1af4:0001 2 (alias eth0) usage is in:0.00% (7058.67bit/s) out:0.00% (5603.67bit/s) | 'Device 1af4:0001 2_usage_in'=0%;80;90;0;100 'Device 1af4:0001 2_usage_out'=0%;80;90;0;100 'Device 1af4:0001 2_traffic_in'=7058.67;0;0;0;0 'Device 1af4:0001         2_traffic_out'=5603.67;0;0;0;0

OK, the important part is over and we can list all server network interfaces plus the usage of a specific network interface. In the next part we will explain the Icinga2 part to add the command and the service to icinga2.


#icinga2 #icinga #nagios #check_nwc_health #network #monitor

How to monitor network cards on Icinga2? (part-2)

Ok for now we have added the plugin to nagios folder and ran some tests on target server's network interfaces. We need to add a command to Icinga2 to use it in service section of Icinga2. To create a new command create a new file in /etc/icinga2/conf.d/commands/check_nwc_command.conf and with the following content:

object CheckCommand "YOUR_COMMAND_NAME" {
     import "plugin-check-command"
     command = [ PluginDir + "/check_nwc_health", "--mode", "interface-usage" ]
     arguments = {
         "-H" = "$address$"
         "-C" = "$community$"
         "--name" = "$int$"
     }
 }

In brief it creates a new command called YOUR_COMMAND_NAME that calls the script check_nwc_health with interface-usage argument to get the bandwidth data.


Now we need to use this command in a service. We have to create a new service which will be used in our hosts configuration sections /etc/icinga2/conf.d/services/if_traffic.conf:

apply Service for (display_name => config in host.vars.int) {
     import "generic-service"
     check_command = "YOUR_COMMAND_NAME"
     vars += config
     assign where host.vars.int
 }

Again in brief the service will be applied on hosts that have a variable section of int in their configuration that we will see a little bit later. YOUR_COMMAND_NAME is the name that we have given in the first part when creating the command.


The final part is to add this service to your desired host. Go to /etc/icinga2/conf.d/hosts and open the file which relates to your host. Host files content start with:

object Host "host-54 (Infra)" {

Add the service like below into your host:

vars.int["YOUR DISPLAY NAME"] = {
     int = "Device 1af4:0001 2"
     community = "YOUR SERVER COMMUNITY STRING"
 }

int is the part that we give the interface name, this should be given from the output of list-interfaces in part-1.

You can go even further like me :) and add these data into Grafana dashboard to have a better understanding of what is happening around you.

#icinga2 #icinga #service #host #command #nagios #interface #network

#grafana #network #network_throughput

nethogs is used to monitor network traffic. You can see which processes use the most bandwidth and hogs the network.

Installtion on debian:

apt-get install nethogs

You can give the nethogs a network interface to see what's going on under the hood:

nethogs eth0

The output would something like:

PID USER       PROGRAM                                                       DEV        SENT      RECEIVED
9023 root      python                                                        eth0        6.083     175.811 KB/sec
20745 root     python                                                        eth0        2.449      45.715 KB/sec
11934 www-da.. nginx: worker process                                         eth0      131.580      20.238 KB/sec
25925 root     /usr/bin/python                                               eth0        3.674      10.090 KB/sec

When nethogs is open, you can press r in order to sort based on RECEIVED or press s to sort based on SENT packets. To change the mode that it is shown for KB/sec press m multiple times and see the output.

#network #sysadmin #linux #nethogs #nethog #network #eth0

How to get list of network interfaces using nagios?

In nagios plugins there is a script called check_nwc_health that can be used to report network interfaces of a remote server. In order to check interface usage of remote server network cards:

cd /usr/lib/nagios/plugins
 ./check_nwc_health --mode interface-usage -H 172.17.131.12 -C YOUR_COMMUNITY_STRING

Give your host address in -H section and your community string in -C param.

#nagios #network #network_usage