Skip to content

Latest commit

 

History

History
196 lines (168 loc) · 11 KB

Report.md

File metadata and controls

196 lines (168 loc) · 11 KB
Raw

CS307 - Assignment 3

Group 10

  • Priyansh Saxena (B14118)
  • Sahil Arora (B14128)
  • Sumant Garg (B14134)
  • Shruti Garg (B13138)

Nagios Overview

Nagios, now known as Nagios Core, is a free and open source computer-software application that monitors systems, networks and infrastructure. Nagios offers monitoring and alerting services for servers, switches, applications and services. It alerts users when things go wrong and alerts them a second time when the problem has been resolved. Currently it provides the following services:

  • Monitoring of network services (SMTP, POP3, HTTP, NNTP, ICMP, SNMP, FTP, SSH).
  • Monitoring of host resources (processor load, disk usage, system logs) on a majority of network operating systems, including Microsoft Windows, using monitoring agents.

What is SNMP ?

  • Simple Network Management Protocol (SNMP) is a popular protocol for network management.
  • It is used for collecting information from, and configuring, network devices, such as servers, printers, hubs, switches, and routers on an Internet Protocol (IP) network.

How to identify units of measurement from SNMP ?

  • To get the metric-descriptor of storage-units on students.iitmandi.ac.in, issue the following command: 
    snmpwalk -v 1 -c public 10.8.1.3 .1.3.6.1.2.1.25.2.3.1.3
  • Next, get the custom-multiplier for a metric (say, Physical Memory) by issuing the following command: 
    snmpwalk -v1 -c public 10.8.1.3 .1.3.6.1.2.1.25.2.3.1.4.1
  • Replace the 1 at the end of the previous command with the appropriate metric-descriptor to get custom-multipliers for that metric.

Services Monitored

  • Ping-Time (Round Trip Average in miiliseconds)
  • RAM Used (in kB)
  • CPU Load (fractional usage)
  • User Load (number of users)

Time-slots used for data-collection

  • Thursday-Friday - 20:00 - 02:00
  • Friday-Saturday - 20:00 - 02:00
  • Sunday-Monday - 20:00 - 02:00

General Observations

  • SNMP is not installed on webopac.iitmandi.ac.in. This shows the following message on RAM Used and User Load in Nagios Dashboard: 
    CRITICAL - Plugin timed out while executing system call
  • User Load and CPU Load show similar trends. This indicates that user-based services are more prominent on all the servers.
  • Packets lost show trends similar to ping-times, denoting network-congestion at times when packets are lost in significant amount.
  • Packet-losses are almost zero for almost the whole day. Week-long average is 1% for all the servers.
  • On Sunday, due to power-failure in G2 building from 20:00 to 21:30, the graph is fairly constant because of missing values and therefore extrapolation. Also, the ping-times are considerably higher.

insite.iitmandi.ac.in

  • RAM Usage of insite unusually increases to around thrice its value in the morning, starting at aroung 7:30 am and continues to remain high for 3-4 hours at a stretch. This may be attributed to some periodic tasks scheduled to run at this time. It was quite low from 18:00 to 6:00 on weekends, owing to people being engaged by Exodia.
  • Ping time is higher than average during evening-hours, from 18:00 to 00:00. This is because insite usage is quite less during class-hours and high during evenings.
  • High ping-time durations closely correspond to high packet-loss durations. Those periods denote network-congestion.
  • User Load goes slightly up after 22:00, and remains a little less from 20:00 to 22:00. This corresponds to the dinner-times.

students.iitmandi.ac.in

  • User Load, and correspondingly CPU Load is low during early-morning hours, from 04:00 to 06:00. This is the time when maximum number of students sleep.
  • CPU load is higher than usual from 12:00 - 02:00 and 06:30 - 08:00 on all days. The first duration can be explained by people using their smart-phones after some deadline or completion of events during weekends, thereby enabling wi-fi and synchronising their e-mail clients. Everyone turns off wi-fi in their smart-phones at night, so the morning-peak may occur because of a similar reason. This also explains the short-peaks at around lunch-time, 12:00 - 14:00.
  • RAM Usage is higher on Saturday and Sunday, whereas User Load and CPU Load is higher on Friday.
  • Packet losses are quite less, 14% at the max.

network.iitmandi.ac.in

  • Generally, CPU load hikes at some intervals every day
  • RAM usage is constant throughout the day on weekends, at a value of around 5.0GB. However, the usage increases to about 6.0GB on Sunday and Monday, corresponding to return to normal routine from fest-time. This is because the services running on this server (like Nagios) are constantly working, and are affected much less by external factors.
  • RAM usage increases slightly over the weekend, as more users start monitoring this server using Nagios.
  • Ping is high in the evening on Friday and very low in the morning, because of the schedule of Exodia on Friday and no-instruction-day. There is a sharp peak after 16:30 in ping-time, user-load and CPU-load.
  • CPU-load and User-load are very low from 18:30 to 00:00 on Saturday, due to students being present in Exodia.
  • User-load is low, and probably this value reflects the number of testers using Nagios to monitor it.

webopac.iitmandi.ac.in

  • There is heavy packet-loss for WebOPAC as compared to other servers.
  • Ping-times are higher than any other server.
  • Due to unavailability of SNMP on this server, there is no data for CPU Load, User Load and RAM Used.

Graphs

  • Day 1: Thursday

    • students.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • insite.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • network.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • webopac.iitmandi.ac.in
    Ping Time Packets Lost

  • Day 2: Friday

    • students.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • insite.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • network.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • webopac.iitmandi.ac.in
    Ping Time Packets Lost

  • Day 3: Sunday

    • students.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • insite.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • network.iitmandi.ac.in
    CPU Load RAM Used
    Ping Time User Load
    • webopac.iitmandi.ac.in
    Ping Time Packets Lost

NRPE Script Run

The machine that is used for calling the script is referred to as master (10.8.17.141) while the one that runs the script is called slave (10.8.15.131).

Following steps were taken to run script by master that is present on slave machine, using NRPE Plugin:

  • The script is put in /usr/local/nagios/libexec in the slave machine. Let the script be test.sh.
  • In the slave machine, we edit the /usr/local/nagios/etc/nrpe.cfg file and include a command for running test.sh by adding the following line: 
    command[testScript]=/usr/local/nagios/libexec/test.sh
  • Edit the file /etc/xinetd.d/nrpe. The only_from line should look like: 
    only_from       = 127.0.0.1 10.8.17.141
  • Edit the file /etc/nagios/nrpe.cfg:
    • Add the IP Address of the master to allowed_hosts, comma-separated list.
    • Add the following line to add the new command:
    command[testScript]=/usr/local/nagios/libexec/test.sh
  • Restart nagios-nrpe-server in the slave: 
    sudo service nagios-nrpe-server restart
  • In master machine, edit the /usr/local/nagios/etc/objects/commands.cfg to include the following command: 
    define command {
    command_name check_nrpe
    command_line $USER1$/check_nrpe -H $HOSTADDRESS$ -c $ARG1$
}
  • The script can now be run by the following command: 
    /usr/local/nagios/libexec/check_nrpe -H 10.8.15.131 -c testScript

References