NAME ¶

 *** Probes ***
 +FPing
 binary = /usr/bin/fping

Here we have just one probe, fping, pinging four hosts.

The fping probe is using the default parameters, some of them supplied from the Database section ("step" and "pings"), and some of them by the probe module.

 *** Targets ***
 probe = FPing
 
 menu = Top
 title = Network Latency Grapher
 remark = Welcome to this SmokePing website.
 
 + mysite1
 menu = Site 1
 title = Hosts in Site 1
 
 ++ myhost1
 host = myhost1.mysite1.example
 ++ myhost2
 host = myhost2.mysite1.example
 
 + mysite2
 menu = Site 2
 title = Hosts in Site 2
 
 ++ myhost3
 host = myhost3.mysite2.example
 ++ myhost4
 host = myhost4.mysite2.example

The hosts are located in two sites of two hosts each, and the configuration has been divided to site sections ('+') and host subsections ('++') accordingly.

Example 2: config.multiple-probes¶

 *** Probes ***
 + FPing
 binary = /usr/bin/fping
 packetsize = 1000
 
 + DNS
 binary = /usr/bin/dig
 lookup = name.example
 pings = 5
 step = 180
 
 + EchoPingHttp
 pings = 5
 url = /test-url

Here we have three probes: FPing for the regular ICMP pings, DNS for name server latency measurement and EchoPingHttp for web servers.

The FPing probe runs with the default parameters, except that the ICMP packet size is 1000 bytes instead of the default 56 bytes.

The DNS and EchoPingHttp probes have been configured to be a bit more gentle with the servers, as they only do 5 queries (pings) instead of the default 20 (or whatever is specified in the Database section). However, DNS queries are made more often: 5 queries every 3 minutes instead of every 5 minutes.

 *** Targets ***
 probe = FPing
 menu = Top
 title = Network Latency Grapher
 remark = Welcome to this SmokePing website.
 
 + network
 menu = Net latency
 title = Network latency (ICMP pings)
 
 ++ myhost1
 host = myhost1.example
 ++ myhost2
 host = myhost2.example
 
 + services
 menu = Service latency
 title = Service latency (DNS, HTTP)
 
 ++ DNS
 probe = DNS
 menu = DNS latency
 title = Service latency (DNS)
 
 +++ dns1
 host = dns1.example
 
 +++ dns2
 host = dns2.example
 
 ++ HTTP
 menu = HTTP latency
 title = Service latency (HTTP)
 
 +++ www1
 host = www1.example
 
 +++ www2
 host = www2.example

The target tree has been divided by the probe used. This does not have to be the case: every target (sub)section can use a different probe, and the same probe can be used in different parts of the config tree.

Example 3: config.fping-instances¶

 *** Probes ***
 + FPing
 binary = /usr/bin/fping
 
 ++ FPingNormal
 offset = 0%
 
 ++ FPingLarge
 packetsize = 5000
 offset = 50%

This example demonstrates the concept of probe instances. The FPingLarge and FPingNormal probes are independent of each other, they just use the same module, FPing. FPingNormal uses the default parameters, and so does FPingLarge except for the 5 kilobyte packetsize. Both use the same fping binary, and its path is configured FPing top section.

The 'offset' parameters make sure the probes don't run at the same time - FPingNormal is run every 'full' 5 minutes (eg. 8:00, 8:05, 8:10 and so on, in wallclock time) while FPingLarge is run halfway through these intervals (eg. 8:02:30, 8:07:30 etc.)

The top FPing section does not define a probe in itself because it has subsections. If we really wanted to have one probe named "FPing", we could do so by making a subsection by that name.

 *** Targets ***
 probe = FPingNormal
 menu = Top
 title = Network Latency Grapher
 remark = Welcome to this SmokePing website.
 
 + network
 menu = Net latency
 title = Network latency (ICMP pings)
 
 ++ myhost1
 menu = myhost1
 title = ICMP latency for myhost1
 
 +++ normal
 title = Normal packetsize (56 bytes)
 probe = FPingNormal
 host = myhost1.example
 
 +++ large
 title = Large packetsize (5000 bytes)
 probe = FPingLarge
 host = myhost1.example
 
 ++ myhost2
 menu = myhost2
 title = ICMP latency for myhost2
 
 +++ normal
 title = Normal packetsize (56 bytes)
 probe = FPingNormal
 host = myhost2.example
 
 +++ large
 title = Large packetsize (5000 bytes)
 probe = FPingLarge
 host = myhost2.example

The target section shows two host, myhost1.example and myhost2.example, being pinged with two differently sized ICMP packets. This time the tree is divided by the target host rather than the probe.

Example 4: config.targetvars-with-Curl¶

 *** Probes ***
 + Curl
 # probe-specific variables
 binary = /usr/bin/curl
 step = 60
 
 # a default for this target-specific variable
 urlformat = http://%host%/

This example explains the difference between probe- and target-specific variables. We use the Curl probe for this.

Every probe supports at least some probe-specific variables. The values of these variables are common to all the targets of the probe, and they can only be configured in the Probes section. In this case, the probe-specific variables are "binary" and "step".

Target-specific variables are supported by most probes, the most notable exception being the FPing probe and its derivatives. Target-specific variables can have different values for different targets. They can be configured in both Probes and Targets sections. The values assigned in the Probes section function become default values that can be overridden in the Targets section.

The documentation of each probe states which of its variables are probe-specific and which are target-specific.

In this case the "urlformat" variable is a target-specific one. It is also quite uncommon, because it can contain a placeholder for the "host" variable in the Targets section. This is not a general feature, its usage is only limited to the "urlformat" variable and the "%host%" escape.

(The reason why the FPing probe does not support target-specific variables is simply the fact that the fping program measures all its targets in one go, so they all have the same parameters. The other probes ping their targets one at a time.)

 *** Targets ***
 probe = Curl
 menu = Top
 title = Network Latency Grapher
 remark = Welcome to this SmokePing website.
 
 + HTTP
 menu = http
 title = HTTP latency 
 
 ++ myhost1
 menu = myhost1
 title = HTTP latency for myhost1
 host = myhost1.example
 
 ++ myhost2
 menu = myhost2
 title = HTTP latency for myhost2
 host = myhost2.example
 
 ++ myhost3
 menu = myhost3
 title = HTTP latency for myhost3 (port 8080!)
 host = myhost3.example
 urlformat = http://%host%:8080/
 
 + FTP
 menu = ftp
 title = FTP latency
 urlformat = ftp://%host%/
 
 ++ myhost1
 menu = myhost1
 title = FTP latency for myhost1
 host = myhost1.example
 
 ++ myhost2
 menu = myhost2
 title = FTP latency for myhost2
 host = myhost2.example

The target tree is divided into an HTTP branch and an FTP one. The servers "myhost1.example" and "myhost2.example" are probed in both. The third server, "myhost3.example", only has an HTTP server, and it's in a non-standard port (8080).

The "urlformat" variable is specified for the whole FTP branch as "ftp://%host%/". For the HTTP branch, the default from the Probes section is used, except for myhost3, which overrides it to tag the port number into the URL.

The myhost3 assignment could just as well have included the hostname verbatim (ie. urlformat = http://myhost3.example:8080/) instead of using the %host% placeholder, but the host variable would still have been required (even though it wouldn't have been used for anything).

Example 5: config.echoping¶

 *** Probes ***
 + FPing
 binary = /usr/bin/fping
 
 # these expect to find echoping in /usr/bin
 # if not, you'll have to specify the location separately for each probe
 # + EchoPing         # uses TCP or UDP echo (port 7)
 # + EchoPingDiscard  # uses TCP or UDP discard (port 9)
 # + EchoPingChargen  # uses TCP chargen (port 19)
 + EchoPingSmtp       # SMTP (25/tcp) for mail servers
 + EchoPingHttps      # HTTPS (443/tcp) for web servers
 + EchoPingHttp       # HTTP (80/tcp) for web servers and caches
 + EchoPingIcp        # ICP (3130/udp) for caches
 # these need at least echoping 6 with the corresponding plugins
 + EchoPingDNS        # DNS (53/udp or tcp) servers
 + EchoPingLDAP       # LDAP (389/tcp) servers
 + EchoPingWhois      # Whois (43/tcp) servers

This example shows most of the echoping-derived probes in action.

 *** Targets ***
 # default probe
 probe = FPing
 
 menu = Top
 title = Network Latency Grapher
 remark = Welcome to this SmokePing website.
 
 + MyServers
 
 menu = My Servers
 title = My Servers 
 
 ++ www-server
 menu = www-server
 title = Web Server (www-server) / ICMP
 # probe = FPing propagated from top
 host = www-server.example
 
 +++ http
 menu = http
 title = Web Server (www-server) / HTTP
 probe = EchoPingHttp
 host = www-server.example 
 # default url is /
 
 +++ https
 menu = https
 title = Web Server (www-server) / HTTPS
 probe = EchoPingHttps
 host = www-server.example
 
 ++ cache
 menu = www-cache
 title = Web Cache (www-cache) / ICMP
 host = www-cache.example
 
 +++ http
 menu = http
 title = www-cache / HTTP
 probe = EchoPingHttp
 host = www-cache.example
 port = 8080 # use the squid port
 url = http://www.somehost.example/
 
 +++ icp
 menu = icp
 title = www-cache / ICP
 probe = EchoPingIcp
 host = www-cache.example
 url = http://www.somehost.example/
 
 ++ mail
 menu = mail-server
 title = Mail Server (mail-server) / ICMP
 host = mail-server.example
 
 +++ smtp
 menu = mail-server / SMTP
 title = Mail Server (mail-server) / SMTP
 probe = EchoPingSmtp
 host = mail-server.example
 
 ++ ldap-server
 menu = ldap-server
 title = ldap-server / ICMP
 host = ldap-server.example
 
 +++ ldap
 menu = ldap-server / LDAP
 title = LDAP Server (ldap-server) / LDAP
 probe = EchoPingLDAP
 ldap_request = (objectclass=*)
 host = ldap-server.example
 
 ++ name-server
 menu = name-server
 title = name-server / ICMP
 host = name-server.example
 
 +++ DNS
 menu = name-server / DNS
 title = DNS Server (name-server) / DNS
 probe = EchoPingDNS
 dns_request = name.example
 host = name-server.example
 
 ++ whois-server
 menu = whois-server
 title = whois-server / ICMP
 host = whois-server.example
 
 +++ Whois
 menu = whois-server / Whois
 title = Whois Server (whois-server) / Whois
 probe = EchoPingWhois
 whois_request = domain.example
 host = whois-server.example

All the servers are pinged both with ICMP (the FPing probe) and their respective echoping probe. The proxy server, www-cache, is probed with both HTTP requests and ICP requests for the same URL.

Example 6: config.template¶

 *** Probes ***
 
 + FPing
 
 binary = /usr/bin/fping
 
 *** Slaves ***
 secrets=/etc/smokeping/smokeping_secrets
 #+boomer
 #display_name=boomer
 #color=0000ff
 
 #+slave2
 #display_name=another
 #color=00ff00

This is the template configuration file distributed with Smokeping. It is included in the examples as well for the sake of completeness.

 *** Targets ***
 
 probe = FPing
 
 menu = Top
 title = Network Latency Grapher
 remark = Welcome to the SmokePing website of xxx Company. \
          Here you will learn all about the latency of our network.
 
 + Local
 
 menu = Local
 title = Local Network
 #parents = owner:/Test/James location:/
 
 ++ LocalMachine
 
 menu = Local Machine
 title = This host
 host = localhost
 #alerts = someloss
 #slaves = boomer slave2