FireBrick FB9000

FB9000 - Ping

The FB9000 is designed to ping things. It can ping lots of hosts (IPv4 and IPv6) and create CQM graphs showing loss and latency, and jitter. Typical applications include a broadband ISP monitoring lines for problems as well as central monitoring of equipment and servers and network links.



  • Tens of thousands of endpoints pinged every second and loss/latency recorded on CQM graphs which can be archived daily.
  • IPv4 and IPv6 addresses can be pinged.
  • Ping data available as svg/png images as well as xml and csv.
  • FireBrick can poll a web server for a list of IPs to ping.
  • Very low power consumption, dual PSU, 1U box. Saves money on space and power in data centre.
  • Easy to see at a glance packet loss, latency and jitter of a remote host.

Data recorded

Loss & Latency

For each ping sent, the loss and latency is recorded. Latency is recorded as min/max/average for each 100 second sample. png and svg image files are created and XML files are also produced which will have a line per 100 pings as follows:


The xsd for this file is also published.


The FireBrick can also assign scores to each 100 second sample. The basis of the score can be customised - eg weighted towards loss or towards latency. This can be used when post-processing the data.

Graph colours

When downloading graphs, various options can be given to customise the style and colours.


We're happy to discuss individual customer requirements with respect to managing ping data etc, please do get in touch.

FB9000 Examples-ping

Generating and Archiving

Creating pings

Ping graphs can be created in a few ways:

  • Defining individual ping addresses in the config
  • Specifying a URL of where they FireBrick can download a list of endpoints.
  • Using the web interface's ping form (Graphs/Ping) to specify individual addresses to ping
  • Making automated HTTP/HTTPS requests to specify individual addresses to ping

Downloading ping data

Graphs and xml data can be viewed and downloaded via the FireBrick's web interface. However, downloading a large number of data items is better done via scripts to retrieve the data you want. You can then store the images offline or post-process the xml data as required.

The manual has more details, but by generating URLs you are able to bulk download the data as well as specify specific options for the data you are downloading, such as date/time ranges, text overlay and colours and other parameters.

Case Study


Yes, the FB2900 model can also generate ping graphs, but for a lower number of endpoints.

Ping is a network utility to see if a remote host responds. By pinging a host you are able to determine the latency (the time it takes to get there and back) and if there is packet loss or jitter. An actual ping packet is an ICMP (internet Control Message Protocol) or ICMPv6 (for IPv6 hosts) packet. The remote host sends back a a reply, and calculation on the timings can be made to measure loss and latency.

Both smokeping and FireBrick's ping graphs show loss and latency. There are some differences though:

  • FireBrick will ping once a second, Smokeping typically sends bursts of pings at longer intervals
  • FireBrick can probably ping more hosts than a single Smokeping instance

Can a FireBrick ping a million hosts? A single FB9000 can ping up to 100,000 hosts, so you'd need 10 units to ping a million!

The payload size can be defined, so you can have multiple entries for a host, specifying different payload sizes.


Further Information


Full features for one price, no per session licencing or other extra costs.


Five year warranty on hardware against any manufacturing defect. Normal working hours / courier replacement. Recommended that two units are used together to provide hardware redundancy. On-going maintenance contracts available for extended hardware support beyond one year.


1U, dual AC 120/240V inlets (monitored), 2 internal fans (monitored), approx 30W total power consumption.

Multi-position 19" rack mount ears, with variation for hanging mount in shallow depth telco racks.

Software upgrades

Free of charge, beta and released software. Internal flash holds up to 8 versions with automatic fallback on crash/watchdog.

Reboot or software upgrade with clean shutdown of L2TP, BGP, VRRP, etc, for minimal disruption. Boot time under 1 second.

UK based s/w support team - email and irc support during office hours.


Configuration defined by an XML document according to a published XSD schema. The configuration may be uploaded and downloaded by HTTP (e.g. using curl). In addition the web interface contains an interactive configuration editor.

Configuration changes are applied as seamlessly as possible when loaded without the need to re-boot.

Command line interface

The command line provides a number of commands to provide viewing of BGP, and pinging status data, as well as clearing BGP sessions. Includes tab completion and interactive help text.


Two 10G SFP+ ports and eight 1G SFP ports allowing 4096 VLANs on each. 100 independent routing tables which can be used with BGP. Each port/VLAN can be attached to a specific routing table.

Access control

Access lists of telnet, ssh, tftp, web, snmp. These can also be attached to an independent routing table for specific port/VLANs.


Syslog to external server with various levels of debugging data available. Logs also available live via command line interface.


CQM (Constant Quality Monitoring( provides graphs for last day, and for nightly archive.

  • Up to 100,000 separate graphs
  • LCP echo every second on every session aggregated to 100 second samples over last day.
  • Packet loss to 1%.
  • Minimum, Average, and Maximum latency to 4 decimal places of ns.
  • Directly http served CSV for analysis.
  • Directly https served PNG/SVG graphs for direct integration in to support systems.
  • Hashed URL for linking for external authenticated viewing.
  • Configurable colours and text and data selection.
  • Configurable scoring of graphs for matching similar lines and identifying common problems.


SNMP (read only) support for a number of functions including interface stats for each port/VLAN in use and individual ping state (up/down).


Simple NTP client to set clock for accurate logging with fallback via list of configured servers.


DHCP client mode available, multiple instances. Also RA client for IPv6 addressing.

RA server for passive IPv6 adress allocation to LAN.


IPv4 VRRP2 and IPv4/6 VRRP3 server.

  • Multiple VRRP IP addresses per port/VLAN.
  • Can use standard floating MAC address, or can use fixed per machine MAC with promiscuous ARPs as configured.
  • Dynamic VRRP priority based on routability of a list of addresses, allows VRRP to only become master when external routing in place.
  • Pingable VRRP addresses for easier diagnostics.

Sales & Dealer Enquiries

phone 01344 400 500
Mon - Fri, 9am-5pm,
calls are recorded
sms 01344 400 500

Support Contact

phone 01344 400 500
Mon-Fri 9am-5pm,
calls are recorded
sms 01344 400 500