riemann-monitors 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/Gemfile +2 -0
- data/LICENSE +21 -0
- data/README.adoc +75 -0
- data/Rakefile +125 -0
- data/bin/riemann-apache-status +98 -0
- data/bin/riemann-bench +71 -0
- data/bin/riemann-cloudant +58 -0
- data/bin/riemann-consul +106 -0
- data/bin/riemann-dir-files-count +55 -0
- data/bin/riemann-dir-space +55 -0
- data/bin/riemann-diskstats +95 -0
- data/bin/riemann-fd +66 -0
- data/bin/riemann-freeswitch +119 -0
- data/bin/riemann-haproxy +58 -0
- data/bin/riemann-health +289 -0
- data/bin/riemann-httpstatus +73 -0
- data/bin/riemann-kvminstance +22 -0
- data/bin/riemann-memcached +38 -0
- data/bin/riemann-net +81 -0
- data/bin/riemann-nginx-status +84 -0
- data/bin/riemann-ntp +35 -0
- data/bin/riemann-proc +131 -0
- data/bin/riemann-varnish +54 -0
- data/bin/riemann-zookeeper +41 -0
- data/data/statfields +49 -0
- data/lib/riemann-monitors/main.rb +111 -0
- data/lib/riemann-monitors/version.rb +6 -0
- data/lib/riemann-monitors.rb +8 -0
- data/project.yaml +12 -0
- data/riemann-monitors.gemspec +73 -0
- metadata +210 -0
@@ -0,0 +1,95 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
require_relative "../lib/riemann-monitors"
|
4
|
+
|
5
|
+
class Riemann::Monitors::Diskstats
|
6
|
+
include Riemann::Monitors
|
7
|
+
|
8
|
+
opt :devices, "Devices to monitor", :type => :strings, :default => nil
|
9
|
+
opt :ignore_devices, "Devices to ignore", :type => :strings, :default =>nil
|
10
|
+
|
11
|
+
def initialize
|
12
|
+
@old_state = nil
|
13
|
+
end
|
14
|
+
|
15
|
+
def state
|
16
|
+
f = File.read('/proc/diskstats')
|
17
|
+
state = f.split("\n").reject { |d| d =~ /(ram|loop)/ }.inject({}) do |s, line|
|
18
|
+
if line =~ /^(?:\s+\d+){2}\s+([\w\d\-]+) (.*)$/
|
19
|
+
dev = $1
|
20
|
+
|
21
|
+
['reads reqs',
|
22
|
+
'reads merged',
|
23
|
+
'reads sector',
|
24
|
+
'reads time',
|
25
|
+
'writes reqs',
|
26
|
+
'writes merged',
|
27
|
+
'writes sector',
|
28
|
+
'writes time',
|
29
|
+
'io reqs',
|
30
|
+
'io time',
|
31
|
+
'io weighted'
|
32
|
+
].map do |service|
|
33
|
+
"#{dev} #{service}"
|
34
|
+
end.zip(
|
35
|
+
$2.split(/\s+/).map { |str| str.to_i }
|
36
|
+
).each do |service, value|
|
37
|
+
s[service] = value
|
38
|
+
end
|
39
|
+
end
|
40
|
+
|
41
|
+
s
|
42
|
+
end
|
43
|
+
|
44
|
+
# Filter interfaces
|
45
|
+
if is = opts[:devices]
|
46
|
+
state = state.select do |service, value|
|
47
|
+
is.include? service.split(' ').first
|
48
|
+
end
|
49
|
+
end
|
50
|
+
|
51
|
+
if ign = opts[:ignore_devices]
|
52
|
+
state = state.reject do |service, value|
|
53
|
+
ign.include? service.split(' ').first
|
54
|
+
end
|
55
|
+
end
|
56
|
+
|
57
|
+
state
|
58
|
+
end
|
59
|
+
|
60
|
+
def tick
|
61
|
+
state = self.state
|
62
|
+
|
63
|
+
if @old_state
|
64
|
+
state.each do |service, metric|
|
65
|
+
|
66
|
+
if service =~ /io reqs$/
|
67
|
+
report(
|
68
|
+
:service => "diskstats " + service,
|
69
|
+
:metric => metric,
|
70
|
+
:state => "ok"
|
71
|
+
)
|
72
|
+
else
|
73
|
+
delta = metric - @old_state[service]
|
74
|
+
|
75
|
+
report(
|
76
|
+
:service => "diskstats " + service,
|
77
|
+
:metric => (delta.to_f / opts[:interval]),
|
78
|
+
:state => "ok"
|
79
|
+
)
|
80
|
+
end
|
81
|
+
|
82
|
+
if service =~ /io time$/
|
83
|
+
report(:service => "diskstats " + service.gsub(/time/, 'util'),
|
84
|
+
:metric => (delta.to_f / (opts[:interval]*1000)),
|
85
|
+
:state => "ok")
|
86
|
+
end
|
87
|
+
end
|
88
|
+
end
|
89
|
+
|
90
|
+
@old_state = state
|
91
|
+
end
|
92
|
+
|
93
|
+
end
|
94
|
+
|
95
|
+
Riemann::Monitors::Diskstats.run
|
data/bin/riemann-fd
ADDED
@@ -0,0 +1,66 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
# Reports current file descriptor use to riemann.
|
4
|
+
# By default reports the total system fd usage, can also report usage of individual processes
|
5
|
+
|
6
|
+
require_relative "../lib/riemann-monitors"
|
7
|
+
|
8
|
+
class Riemann::Monitors::Health
|
9
|
+
include Riemann::Monitors
|
10
|
+
|
11
|
+
opt :fd_sys_warning, "open file descriptor threshold for system", :default => 800
|
12
|
+
opt :fd_sys_critical, "open file descriptor critical threshold for system", :default => 900
|
13
|
+
opt :fd_proc_warning, "open file descriptor threshold for process", :default => 800
|
14
|
+
opt :fd_proc_critical, "open file descriptor critical threshold for process", :default => 900
|
15
|
+
opt :processes, "list of processes to measure fd usage in addition to system total", :type => :ints
|
16
|
+
|
17
|
+
def initialize
|
18
|
+
@limits = {
|
19
|
+
:fd => {:critical => opts[:fd_sys_critical], :warning => opts[:fd_sys_warning]},
|
20
|
+
:process => {:critical => opts[:fd_proc_critical], :warning => opts[:fd_proc_warning]},
|
21
|
+
}
|
22
|
+
ostype = `uname -s`.chomp.downcase
|
23
|
+
puts "WARNING: OS '#{ostype}' not explicitly supported. Falling back to Linux" unless ostype == "linux"
|
24
|
+
@fd = method :linux_fd
|
25
|
+
end
|
26
|
+
|
27
|
+
def alert(service, state, metric, description)
|
28
|
+
report(
|
29
|
+
:service => service.to_s,
|
30
|
+
:state => state.to_s,
|
31
|
+
:metric => metric.to_f,
|
32
|
+
:description => description
|
33
|
+
)
|
34
|
+
end
|
35
|
+
|
36
|
+
def linux_fd
|
37
|
+
sys_used = Integer(`lsof | wc -l`)
|
38
|
+
if sys_used > @limits[:fd][:critical]
|
39
|
+
alert "fd sys", :critical, sys_used, "system is using #{sys_used} fds"
|
40
|
+
elsif sys_used > @limits[:fd][:warning]
|
41
|
+
alert "fd sys", :warning, sys_used, "system is using #{sys_used} fds"
|
42
|
+
else
|
43
|
+
alert "fd sys", :ok, sys_used, "system is using #{sys_used} fds"
|
44
|
+
end
|
45
|
+
|
46
|
+
unless opts[:processes].nil?
|
47
|
+
opts[:processes].each do |process|
|
48
|
+
used = Integer(`lsof -p #{process} | wc -l`)
|
49
|
+
name, pid = `ps axo comm,pid | grep -w #{process}`.split
|
50
|
+
if used > @limits[:process][:critical]
|
51
|
+
alert "fd #{name} #{process}", :critical, used, "process #{name} #{process} is using #{used} fds"
|
52
|
+
elsif used > @limits[:process][:warning]
|
53
|
+
alert "fd #{name} #{process}", :warning, used, "process #{name} #{process} is using #{used} fds"
|
54
|
+
else
|
55
|
+
alert "fd #{name} #{process}", :ok, used, "process #{name} #{process} is using #{used} fds"
|
56
|
+
end
|
57
|
+
end
|
58
|
+
end
|
59
|
+
end
|
60
|
+
|
61
|
+
def tick
|
62
|
+
@fd.call
|
63
|
+
end
|
64
|
+
end
|
65
|
+
|
66
|
+
Riemann::Monitors::Health.run
|
@@ -0,0 +1,119 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
require File.expand_path('../../lib/riemann/tools', __FILE__)
|
4
|
+
|
5
|
+
class Riemann::Monitors::FreeSWITCH
|
6
|
+
include Riemann::Monitors
|
7
|
+
|
8
|
+
opt :calls_warning, "Calls warning threshold", :default => 100
|
9
|
+
opt :calls_critical, "Calls critical threshold", :default => 300
|
10
|
+
opt :pid_file, "FreeSWITCH daemon pidfile", :type => String, :default => "/var/run/freeswitch/freeswitch.pid"
|
11
|
+
|
12
|
+
def initialize
|
13
|
+
@limits = {
|
14
|
+
:calls => {:critical => opts[:calls_critical], :warning => opts[:calls_warning]}
|
15
|
+
}
|
16
|
+
end
|
17
|
+
|
18
|
+
def dead_proc?(pid)
|
19
|
+
begin
|
20
|
+
Process.getpgid(pid)
|
21
|
+
false
|
22
|
+
rescue Errno::ESRCH
|
23
|
+
true
|
24
|
+
end
|
25
|
+
end
|
26
|
+
|
27
|
+
def alert(service, state, metric, description)
|
28
|
+
report(
|
29
|
+
:service => service.to_s,
|
30
|
+
:state => state.to_s,
|
31
|
+
:metric => metric.to_f,
|
32
|
+
:description => description
|
33
|
+
)
|
34
|
+
end
|
35
|
+
|
36
|
+
def exec_with_timeout(cmd, timeout)
|
37
|
+
pid = Process.spawn(cmd, {[:err,:out] => :close, :pgroup => true})
|
38
|
+
begin
|
39
|
+
Timeout.timeout(timeout) do
|
40
|
+
Process.waitpid(pid, 0)
|
41
|
+
$?.exitstatus == 0
|
42
|
+
end
|
43
|
+
rescue Timeout::Error
|
44
|
+
Process.kill(15, -Process.getpgid(pid))
|
45
|
+
puts "Killed pid: #{pid}"
|
46
|
+
false
|
47
|
+
end
|
48
|
+
end
|
49
|
+
|
50
|
+
def tick
|
51
|
+
# Determine how many current calls I have according to FreeSWITCH
|
52
|
+
fs_calls = %x[fs_cli -x "show calls count"| grep -Po '^\\d+'].to_i
|
53
|
+
|
54
|
+
# Determine how many current channels I have according to FreeSWITCH
|
55
|
+
fs_channels = %x[fs_cli -x "show channels count"| grep -Po '^\\d+'].to_i
|
56
|
+
|
57
|
+
# Determine how many conferences I have according to FreeSWITCH
|
58
|
+
fs_conferences = %x[fs_cli -x "conference list"| grep -Pco '^Conference'].to_i
|
59
|
+
|
60
|
+
# Try to read pidfile. If it fails use Devil's dummy PID
|
61
|
+
begin
|
62
|
+
fs_pid = File.read(opts[:pid_file]).to_i
|
63
|
+
rescue
|
64
|
+
puts "Couldn't read pidfile: #{opts[:pid_file]}"
|
65
|
+
fs_pid = -666
|
66
|
+
end
|
67
|
+
|
68
|
+
fs_threads = fs_pid > 0 ? %x[ps huH p #{fs_pid} | wc -l].to_i : 0
|
69
|
+
|
70
|
+
# Submit calls to riemann
|
71
|
+
if fs_calls > @limits[:calls][:critical]
|
72
|
+
alert "FreeSWITCH current calls", :critical, fs_calls, "Number of calls are #{fs_calls}"
|
73
|
+
elsif fs_calls > @limits[:calls][:warning]
|
74
|
+
alert "FreeSWITCH current calls", :warning, fs_calls, "Number of calls are #{fs_calls}"
|
75
|
+
else
|
76
|
+
alert "FreeSWITCH current calls", :ok, fs_calls, "Number of calls are #{fs_calls}"
|
77
|
+
end
|
78
|
+
|
79
|
+
# Submit channels to riemann
|
80
|
+
if fs_channels > @limits[:calls][:critical]
|
81
|
+
alert "FreeSWITCH current channels", :critical, fs_channels, "Number of channels are #{fs_channels}"
|
82
|
+
elsif fs_channels > @limits[:calls][:warning]
|
83
|
+
alert "FreeSWITCH current channels", :warning, fs_channels, "Number of channels are #{fs_channels}"
|
84
|
+
else
|
85
|
+
alert "FreeSWITCH current channels", :ok, fs_channels, "Number of channels are #{fs_channels}"
|
86
|
+
end
|
87
|
+
|
88
|
+
# Submit conferences to riemann
|
89
|
+
if fs_conferences > @limits[:calls][:critical]
|
90
|
+
alert "FreeSWITCH current conferences", :critical, fs_conferences, "Number of conferences are #{fs_conferences}"
|
91
|
+
elsif fs_conferences > @limits[:calls][:warning]
|
92
|
+
alert "FreeSWITCH current conferences", :warning, fs_conferences, "Number of conferences are #{fs_conferences}"
|
93
|
+
else
|
94
|
+
alert "FreeSWITCH current conferences", :ok, fs_conferences, "Number of conferences are #{fs_conferences}"
|
95
|
+
end
|
96
|
+
|
97
|
+
# Submit threads to riemann
|
98
|
+
if fs_threads
|
99
|
+
alert "FreeSWITCH current threads", :ok, fs_threads, "Number of threads are #{fs_threads}"
|
100
|
+
end
|
101
|
+
|
102
|
+
# Submit status to riemann
|
103
|
+
if dead_proc?(fs_pid)
|
104
|
+
alert "FreeSWITCH status", :critical, -1, "FreeSWITCH service status: not running"
|
105
|
+
else
|
106
|
+
alert "FreeSWITCH status", :ok, nil, "FreeSWITCH service status: running"
|
107
|
+
end
|
108
|
+
|
109
|
+
# Submit CLI status to riemann using timeout in case it's unresponsive
|
110
|
+
if exec_with_timeout("fs_cli -x status", 2)
|
111
|
+
alert "FreeSWITCH CLI status", :ok, nil, "FreeSWITCH CLI status: responsive"
|
112
|
+
else
|
113
|
+
alert "FreeSWITCH CLI status", :critical, -1, "FreeSWITCH CLI status: not responding"
|
114
|
+
end
|
115
|
+
|
116
|
+
end
|
117
|
+
end
|
118
|
+
|
119
|
+
Riemann::Monitors::FreeSWITCH.run
|
data/bin/riemann-haproxy
ADDED
@@ -0,0 +1,58 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
# Gathers haproxy CSV statistics and submits them to Riemann.
|
4
|
+
|
5
|
+
require File.expand_path('../../lib/riemann/tools', __FILE__)
|
6
|
+
|
7
|
+
class Riemann::Monitors::Haproxy
|
8
|
+
include Riemann::Monitors
|
9
|
+
require 'net/http'
|
10
|
+
require 'csv'
|
11
|
+
|
12
|
+
opt :stats_url, "Full url to haproxy stats (eg: https://user:password@host.com:9999/stats)", :required => true, :type => :string
|
13
|
+
|
14
|
+
def initialize
|
15
|
+
@uri = URI(opts[:stats_url]+';csv')
|
16
|
+
end
|
17
|
+
|
18
|
+
def tick
|
19
|
+
csv = CSV.parse(get_csv.body.split("# ")[1], { :headers => true })
|
20
|
+
csv.each do |row|
|
21
|
+
row = row.to_hash
|
22
|
+
ns = "haproxy #{row['pxname']} #{row['svname']}"
|
23
|
+
row.each do |property, metric|
|
24
|
+
unless (property.nil? || property == 'pxname' || property == 'svname')
|
25
|
+
report(
|
26
|
+
:host => @uri.host,
|
27
|
+
:service => "#{ns} #{property}",
|
28
|
+
:metric => metric.to_f,
|
29
|
+
:tags => ['haproxy']
|
30
|
+
)
|
31
|
+
end
|
32
|
+
end
|
33
|
+
|
34
|
+
report(
|
35
|
+
:host => @uri.host,
|
36
|
+
:service => "#{ns} state",
|
37
|
+
:state => (['UP', 'OPEN'].include?(row['status']) ? 'ok' : 'critical'),
|
38
|
+
:tags => ['haproxy']
|
39
|
+
)
|
40
|
+
end
|
41
|
+
end
|
42
|
+
|
43
|
+
def get_csv
|
44
|
+
http = Net::HTTP.new(@uri.host, @uri.port)
|
45
|
+
http.use_ssl = true if @uri.scheme == 'https'
|
46
|
+
http.start do |h|
|
47
|
+
get = Net::HTTP::Get.new(@uri.request_uri)
|
48
|
+
unless @uri.userinfo.nil?
|
49
|
+
userinfo = @uri.userinfo.split(":")
|
50
|
+
get.basic_auth userinfo[0], userinfo[1]
|
51
|
+
end
|
52
|
+
h.request get
|
53
|
+
end
|
54
|
+
end
|
55
|
+
|
56
|
+
end
|
57
|
+
|
58
|
+
Riemann::Monitors::Haproxy.run
|
data/bin/riemann-health
ADDED
@@ -0,0 +1,289 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
# Reports current CPU, disk, load average, and memory use to riemann.
|
4
|
+
|
5
|
+
require_relative "../lib/riemann-monitors"
|
6
|
+
|
7
|
+
class Riemann::Monitors::Health
|
8
|
+
include Riemann::Monitors
|
9
|
+
|
10
|
+
opt :cpu_warning, "CPU warning threshold (fraction of total jiffies)", :default => 0.9
|
11
|
+
opt :cpu_critical, "CPU critical threshold (fraction of total jiffies)", :default => 0.95
|
12
|
+
opt :disk_warning, "Disk warning threshold (fraction of space used)", :default => 0.9
|
13
|
+
opt :disk_critical, "Disk critical threshold (fraction of space used)", :default => 0.95
|
14
|
+
opt :load_warning, "Load warning threshold (load average / core)", :default => 3
|
15
|
+
opt :load_critical, "Load critical threshold (load average / core)", :default => 8
|
16
|
+
opt :memory_warning, "Memory warning threshold (fraction of RAM)", :default => 0.85
|
17
|
+
opt :memory_critical, "Memory critical threshold (fraction of RAM)", :default => 0.95
|
18
|
+
opt :checks, "A list of checks to run.", :type => :strings, :default => ['cpu', 'load', 'memory', 'disk']
|
19
|
+
|
20
|
+
def initialize
|
21
|
+
@limits = {
|
22
|
+
:cpu => {:critical => opts[:cpu_critical], :warning => opts[:cpu_warning]},
|
23
|
+
:disk => {:critical => opts[:disk_critical], :warning => opts[:disk_warning]},
|
24
|
+
:load => {:critical => opts[:load_critical], :warning => opts[:load_warning]},
|
25
|
+
:memory => {:critical => opts[:memory_critical], :warning => opts[:memory_warning]}
|
26
|
+
}
|
27
|
+
case (@ostype = `uname -s`.chomp.downcase)
|
28
|
+
when 'darwin'
|
29
|
+
@cores = `sysctl -n hw.ncpu`.to_i
|
30
|
+
@cpu = method :darwin_cpu
|
31
|
+
@disk = method :disk
|
32
|
+
@load = method :darwin_load
|
33
|
+
@memory = method :darwin_memory
|
34
|
+
darwin_top
|
35
|
+
when 'freebsd'
|
36
|
+
@cores = `sysctl -n hw.ncpu`.to_i
|
37
|
+
@cpu = method :freebsd_cpu
|
38
|
+
@disk = method :disk
|
39
|
+
@load = method :freebsd_load
|
40
|
+
@memory = method :freebsd_memory
|
41
|
+
else
|
42
|
+
@cores = cores
|
43
|
+
puts "WARNING: OS '#{@ostype}' not explicitly supported. Falling back to Linux" unless @ostype == "linux"
|
44
|
+
@cpu = method :linux_cpu
|
45
|
+
@disk = method :disk
|
46
|
+
@load = method :linux_load
|
47
|
+
@memory = method :linux_memory
|
48
|
+
end
|
49
|
+
|
50
|
+
opts[:checks].each do |check|
|
51
|
+
case check
|
52
|
+
when "disk"
|
53
|
+
@disk_enabled = true
|
54
|
+
when "load"
|
55
|
+
@load_enabled = true
|
56
|
+
when "cpu"
|
57
|
+
@cpu_enabled = true
|
58
|
+
when "memory"
|
59
|
+
@memory_enabled = true
|
60
|
+
end
|
61
|
+
end
|
62
|
+
end
|
63
|
+
|
64
|
+
def alert(service, state, metric, description)
|
65
|
+
report(
|
66
|
+
:service => service.to_s,
|
67
|
+
:state => state.to_s,
|
68
|
+
:metric => metric.to_f,
|
69
|
+
:description => description
|
70
|
+
)
|
71
|
+
end
|
72
|
+
|
73
|
+
def cores
|
74
|
+
i = 0;
|
75
|
+
File.read("/proc/cpuinfo").split(/\n\n/).inject({}) do |cores, p|
|
76
|
+
physical_id = p[/physical id\s+:\s+(\d+)/, 1]
|
77
|
+
core_id = p[/core id\s+:\s+(\d+)/, 1]
|
78
|
+
if physical_id and core_id
|
79
|
+
cores["#{physical_id}:#{core_id}"] = true
|
80
|
+
elsif physical_id
|
81
|
+
cores["#{physical_id}:"] = true
|
82
|
+
else
|
83
|
+
cores[i += 1] = true;
|
84
|
+
end
|
85
|
+
|
86
|
+
cores
|
87
|
+
end.size
|
88
|
+
end
|
89
|
+
|
90
|
+
def report_pct(service, fraction, report)
|
91
|
+
if fraction
|
92
|
+
if fraction > @limits[service][:critical]
|
93
|
+
alert service, :critical, fraction, "#{sprintf("%.2f", fraction * 100)}% #{report}"
|
94
|
+
elsif fraction > @limits[service][:warning]
|
95
|
+
alert service, :warning, fraction, "#{sprintf("%.2f", fraction * 100)}% #{report}"
|
96
|
+
else
|
97
|
+
alert service, :ok, fraction, "#{sprintf("%.2f", fraction * 100)}% #{report}"
|
98
|
+
end
|
99
|
+
end
|
100
|
+
end
|
101
|
+
|
102
|
+
def linux_cpu
|
103
|
+
new = File.read('/proc/stat')
|
104
|
+
unless new[/cpu\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)/]
|
105
|
+
alert 'cpu', :unknown, nil, "/proc/stat doesn't include a CPU line"
|
106
|
+
return false
|
107
|
+
end
|
108
|
+
u2, n2, s2, i2 = [$1, $2, $3, $4].map { |e| e.to_i }
|
109
|
+
|
110
|
+
if @old_cpu
|
111
|
+
u1, n1, s1, i1 = @old_cpu
|
112
|
+
|
113
|
+
used = (u2+n2+s2) - (u1+n1+s1)
|
114
|
+
total = used + i2-i1
|
115
|
+
fraction = used.to_f / total
|
116
|
+
|
117
|
+
report_pct :cpu, fraction, "user+nice+system\n\n#{`ps -eo pcpu,pid,comm | sort -nrb -k1 | head -10`.chomp}"
|
118
|
+
end
|
119
|
+
|
120
|
+
@old_cpu = [u2, n2, s2, i2]
|
121
|
+
end
|
122
|
+
|
123
|
+
def linux_load
|
124
|
+
load = File.read('/proc/loadavg').split(/\s+/)[0].to_f / @cores
|
125
|
+
if load > @limits[:load][:critical]
|
126
|
+
alert "load", :critical, load, "1-minute load average/core is #{load}"
|
127
|
+
elsif load > @limits[:load][:warning]
|
128
|
+
alert "load", :warning, load, "1-minute load average/core is #{load}"
|
129
|
+
else
|
130
|
+
alert "load", :ok, load, "1-minute load average/core is #{load}"
|
131
|
+
end
|
132
|
+
end
|
133
|
+
|
134
|
+
def linux_memory
|
135
|
+
m = File.read('/proc/meminfo').split(/\n/).inject({}) { |info, line|
|
136
|
+
x = line.split(/:?\s+/)
|
137
|
+
# Assume kB...
|
138
|
+
info[x[0]] = x[1].to_i
|
139
|
+
info
|
140
|
+
}
|
141
|
+
|
142
|
+
free = m['MemFree'].to_i + m['Buffers'].to_i + m['Cached'].to_i
|
143
|
+
total = m['MemTotal'].to_i
|
144
|
+
fraction = 1 - (free.to_f / total)
|
145
|
+
|
146
|
+
report_pct :memory, fraction, "used\n\n#{`ps -eo pmem,pid,comm | sort -nrb -k1 | head -10`.chomp}"
|
147
|
+
end
|
148
|
+
|
149
|
+
def freebsd_cpu
|
150
|
+
u2, n2, s2, t2, i2 = `sysctl -n kern.cp_time 2>/dev/null`.split.map{ |e| e.to_i } #FreeBSD has 5 cpu stats
|
151
|
+
|
152
|
+
if @old_cpu
|
153
|
+
u1, n1, s1, t1, i1 = @old_cpu
|
154
|
+
|
155
|
+
used = (u2+n2+s2+t2) - (u1+n1+s1+t1)
|
156
|
+
total = used + i2-i1
|
157
|
+
fraction = used.to_f / total
|
158
|
+
|
159
|
+
report_pct :cpu, fraction, "user+nice+sytem+interrupt\n\n#{`ps -axo pcpu,pid,comm | sort -nrb -k1 | head -10`.chomp}"
|
160
|
+
end
|
161
|
+
|
162
|
+
@old_cpu = [u2, n2, s2, t2, i2]
|
163
|
+
end
|
164
|
+
|
165
|
+
def freebsd_load
|
166
|
+
m = `uptime`.split(':')[-1].chomp.gsub(/\s+/,'').split(',')
|
167
|
+
load = m[0].to_f / @cores
|
168
|
+
if load > @limits[:load][:critical]
|
169
|
+
alert "load", :critical, load, "1-minute load average/core is #{load}"
|
170
|
+
elsif load > @limits[:load][:warning]
|
171
|
+
alert "load", :warning, load, "1-minute load average/core is #{load}"
|
172
|
+
else
|
173
|
+
alert "load", :ok, load, "1-minute load average/core is #{load}"
|
174
|
+
end
|
175
|
+
end
|
176
|
+
|
177
|
+
def freebsd_memory
|
178
|
+
meminfo = `sysctl -n vm.stats.vm.v_page_count vm.stats.vm.v_wire_count vm.stats.vm.v_active_count 2>/dev/null`.chomp.split
|
179
|
+
fraction = (meminfo[1].to_f + meminfo[2].to_f) / meminfo[0].to_f
|
180
|
+
|
181
|
+
report_pct :memory, fraction, "used\n\n#{`ps -axo pmem,pid,comm | sort -nrb -k1 | head -10`.chomp}"
|
182
|
+
end
|
183
|
+
|
184
|
+
def darwin_top
|
185
|
+
raw = `top -l 1 | grep -i "^\\(cpu\\|physmem\\|load\\)"`.chomp
|
186
|
+
@topdata = {:stamp => Time.now.to_i }
|
187
|
+
raw.each_line do |ln|
|
188
|
+
if ln.match(/Load Avg: [0-9.]+, [0-9.]+, ([0-9.])+/i)
|
189
|
+
@topdata[:load] = $1.to_f
|
190
|
+
elsif ln.match(/CPU usage: [0-9.]+% user, [0-9.]+% sys, ([0-9.]+)% idle/i)
|
191
|
+
@topdata[:cpu] = 1 - ($1.to_f / 100)
|
192
|
+
elsif mdat = ln.match(/PhysMem: ([0-9]+)([BKMGT]) wired, ([0-9]+)([BKMGT]) active, ([0-9]+)([BKMGT]) inactive, ([0-9]+)([BKMGT]) used, ([0-9]+)([BKMGT]) free/i)
|
193
|
+
wired = mdat[1].to_i * (1024 ** "BKMGT".index(mdat[2]))
|
194
|
+
active = mdat[3].to_i * (1024 ** "BKMGT".index(mdat[4]))
|
195
|
+
inactive = mdat[5].to_i * (1024 ** "BKMGT".index(mdat[6]))
|
196
|
+
used = mdat[7].to_i * (1024 ** "BKMGT".index(mdat[8]))
|
197
|
+
free = mdat[9].to_i * (1024 ** "BKMGT".index(mdat[10]))
|
198
|
+
@topdata[:memory] = (wired + active + used).to_f / (wired + active + used + inactive + free)
|
199
|
+
# This is for OSX Mavericks which
|
200
|
+
# uses a different format for top
|
201
|
+
# Example: PhysMem: 4662M used (1328M wired), 2782M unused.
|
202
|
+
elsif mdat = ln.match(/PhysMem: ([0-9]+)([BKMGT]) used \(([0-9]+)([BKMGT]) wired\), ([0-9]+)([BKMGT]) unused/i)
|
203
|
+
used = mdat[1].to_i * (1024 ** "BKMGT".index(mdat[2]))
|
204
|
+
wired = mdat[3].to_i * (1024 ** "BKMGT".index(mdat[4]))
|
205
|
+
unused = mdat[5].to_i * (1024 ** "BKMGT".index(mdat[6]))
|
206
|
+
@topdata[:memory] = (used).to_f / (used + unused)
|
207
|
+
end
|
208
|
+
end
|
209
|
+
end
|
210
|
+
|
211
|
+
def darwin_cpu
|
212
|
+
darwin_top unless (Time.now.to_i - @topdata[:stamp]) < opts[:interval]
|
213
|
+
unless @topdata[:cpu]
|
214
|
+
alert 'cpu', :unknown, nil, "unable to get CPU stats from top"
|
215
|
+
return false
|
216
|
+
end
|
217
|
+
report_pct :cpu, @topdata[:cpu], "usage\n\n#{`ps -eo pcpu,pid,comm | sort -nrb -k1 | head -10`.chomp}"
|
218
|
+
end
|
219
|
+
|
220
|
+
def darwin_load
|
221
|
+
darwin_top unless (Time.now.to_i - @topdata[:stamp]) < opts[:interval]
|
222
|
+
unless @topdata[:load]
|
223
|
+
alert 'load', :unknown, nil, "unable to get load ave from top"
|
224
|
+
return false
|
225
|
+
end
|
226
|
+
metric = @topdata[:load] / @cores
|
227
|
+
if metric > @limits[:load][:critical]
|
228
|
+
alert "load", :critical, metric, "1-minute load average per core is #{metric}"
|
229
|
+
elsif metric > @limits[:load][:warning]
|
230
|
+
alert "load", :warning, metric, "1-minute load average per core is #{metric}"
|
231
|
+
else
|
232
|
+
alert "load", :ok, metric, "1-minute load average per core is #{metric}"
|
233
|
+
end
|
234
|
+
end
|
235
|
+
|
236
|
+
def darwin_memory
|
237
|
+
darwin_top unless (Time.now.to_i - @topdata[:stamp]) < opts[:interval]
|
238
|
+
unless @topdata[:memory]
|
239
|
+
alert 'memory', :unknown, nil, "unable to get memory data from top"
|
240
|
+
return false
|
241
|
+
end
|
242
|
+
report_pct :memory, @topdata[:memory], "usage\n\n#{`ps -eo pmem,pid,comm | sort -nrb -k1 | head -10`.chomp}"
|
243
|
+
end
|
244
|
+
|
245
|
+
def df
|
246
|
+
case @ostype
|
247
|
+
when 'darwin', 'freebsd'
|
248
|
+
`df -P -t noiso9660`
|
249
|
+
else
|
250
|
+
`df -P --exclude-type=iso9660`
|
251
|
+
end
|
252
|
+
end
|
253
|
+
|
254
|
+
def disk
|
255
|
+
df.split(/\n/).each do |r|
|
256
|
+
f = r.split(/\s+/)
|
257
|
+
next if f[0] == 'Filesystem'
|
258
|
+
next unless f[0] =~ /\// # Needs at least one slash in the mount path
|
259
|
+
|
260
|
+
# Calculate capacity
|
261
|
+
x = f[4].to_f/100
|
262
|
+
|
263
|
+
if x > @limits[:disk][:critical]
|
264
|
+
alert "disk #{f[5]}", :critical, x, "#{f[4]} used"
|
265
|
+
elsif x > @limits[:disk][:warning]
|
266
|
+
alert "disk #{f[5]}", :warning, x, "#{f[4]} used"
|
267
|
+
else
|
268
|
+
alert "disk #{f[5]}", :ok, x, "#{f[4]} used"
|
269
|
+
end
|
270
|
+
end
|
271
|
+
end
|
272
|
+
|
273
|
+
def tick
|
274
|
+
if @cpu_enabled
|
275
|
+
@cpu.call
|
276
|
+
end
|
277
|
+
if @memory_enabled
|
278
|
+
@memory.call
|
279
|
+
end
|
280
|
+
if @disk_enabled
|
281
|
+
@disk.call
|
282
|
+
end
|
283
|
+
if @load_enabled
|
284
|
+
@load.call
|
285
|
+
end
|
286
|
+
end
|
287
|
+
end
|
288
|
+
|
289
|
+
Riemann::Monitors::Health.run
|
@@ -0,0 +1,73 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
# Checks the status of an HTTP request and submits it to Riemann
|
4
|
+
require_relative "../lib/riemann-monitors"
|
5
|
+
|
6
|
+
class Riemann::Monitors::HTTPStatus
|
7
|
+
include Riemann::Monitors
|
8
|
+
require "net/http"
|
9
|
+
require "uri"
|
10
|
+
|
11
|
+
opt :uri, "Target URI", type: String, default: 'http://localhost/'
|
12
|
+
opt :sitename, "Site name (abbreviation for identifying uri)", type: String
|
13
|
+
opt :method, "HTTP Method (get, head)", type: String, default: "head"
|
14
|
+
|
15
|
+
def initialize
|
16
|
+
@uri = URI.parse(opts[:uri])
|
17
|
+
@sitename = opts[:sitename]&.gsub(/\//, '_') if opts.has_key?(:sitename)
|
18
|
+
@servicename = ["http", opts[:method], @sitename].join("/")
|
19
|
+
@verbclass = case opts[:method]
|
20
|
+
when /\Aget\z/i
|
21
|
+
Net::HTTP::Get
|
22
|
+
when /\Ahead\z/i
|
23
|
+
Net::HTTP::Head
|
24
|
+
else
|
25
|
+
raise "This script only makes sense with get and head methods. You might want a custom script."
|
26
|
+
end
|
27
|
+
end
|
28
|
+
|
29
|
+
def tick
|
30
|
+
response, roundtrip = get_connection()
|
31
|
+
unless (response).nil?
|
32
|
+
metrics = {service: @servicename,
|
33
|
+
description: @sitename,
|
34
|
+
tags: ["http"]}
|
35
|
+
c = response.code&.to_i
|
36
|
+
if !c.nil?
|
37
|
+
metrics[:state] = "ok"
|
38
|
+
metrics[:metric] = c
|
39
|
+
end
|
40
|
+
if !roundtrip.nil?
|
41
|
+
metrics[:roundtrip] = roundtrip
|
42
|
+
end
|
43
|
+
if !response.body.nil? && @verbclass != Net::HTTP::Head
|
44
|
+
metrics[:body_length] = response.body.length
|
45
|
+
end
|
46
|
+
report(metrics)
|
47
|
+
end
|
48
|
+
end
|
49
|
+
|
50
|
+
def get_connection()
|
51
|
+
response = nil
|
52
|
+
roundtrip = nil
|
53
|
+
begin
|
54
|
+
p1 = Time.now
|
55
|
+
Net::HTTP.start(@uri.host, @uri.port) do |http|
|
56
|
+
http.open_timeout = opts[:timeout]
|
57
|
+
request = @verbclass.new(@uri)
|
58
|
+
response = http.request(request)
|
59
|
+
response.read_body
|
60
|
+
end
|
61
|
+
p2 = Time.now
|
62
|
+
roundtrip = p2 - p1
|
63
|
+
rescue => e
|
64
|
+
report(:service => @servicename,
|
65
|
+
:state => "critical",
|
66
|
+
:description => "http connection failure: #{e.class} - #{e.message}",
|
67
|
+
:tags => ["http", "action"])
|
68
|
+
end
|
69
|
+
[response, roundtrip]
|
70
|
+
end
|
71
|
+
end
|
72
|
+
|
73
|
+
Riemann::Monitors::HTTPStatus.run
|