new web_log alarms to account 1xx, 2xx and 304 as successful; fixes #1780, fixes #1781
Costa Tsaousis committed
Feb 13, 2017 at 16:47 UTC
e5207d67cbddaed531b6b49f270a9e2f9f4149bf
1 file changed
+32
-40
conf.d/health.d/web_log.conf
+32
-40
@@ -25,7 +25,7 @@ families: *
25
# i.e. when there are at least 120 requests during the last minute
26
27
template: 1m_requests
28
- on: web_log.response_codes
28
+ on: web_log.response_statuses
29
families: *
30
lookup: sum -1m unaligned
31
calc: ($this == 0)?(1):($this)
@@ -33,57 +33,49 @@ families: *
33
every: 10s
34
info: the sum of all HTTP requests over the last minute
35
36
-template: 1m_2xx
37
- on: web_log.response_codes
38
-families: *
39
- lookup: sum -1m unaligned of 2xx
40
- calc: ($this == 0)?(1):($this)
41
- units: requests
42
- every: 10s
43
- info: the sum of successful HTTP requests over the last minute
44
-
36
template: 1m_successful
46
- on: web_log.response_codes
37
+ on: web_log.response_statuses
38
families: *
48
- calc: $1m_2xx * 100 / $1m_requests
39
+ lookup: sum -1m unaligned of successful_requests
40
+ calc: $this * 100 / $1m_requests
41
units: %
42
every: 10s
51
- warn: ($1m_requests > 120) ? ($this < (($status >= $WARNING ) ? ( 98 ) : ( 95 )) ) : ( 0 )
52
- crit: ($1m_requests > 120) ? ($this < (($status == $CRITICAL) ? ( 95 ) : ( 90 )) ) : ( 0 )
43
+ warn: ($1m_requests > 120) ? ($this < (($status >= $WARNING ) ? ( 95 ) : ( 85 )) ) : ( 0 )
44
+ crit: ($1m_requests > 120) ? ($this < (($status == $CRITICAL) ? ( 85 ) : ( 75 )) ) : ( 0 )
45
delay: down 15m multiplier 1.5 max 1h
54
- info: the ratio of successful HTTP responses (2xx) over the last minute
46
+ info: the ratio of successful HTTP responses (1xx, 2xx, 304) over the last minute
47
to: webmaster
48
49
template: 1m_redirects
58
- on: web_log.detailed_response_codes
50
+ on: web_log.response_statuses
51
families: *
60
- lookup: sum -1m unaligned of 301,303,307,308
52
+ lookup: sum -1m unaligned of redirects
53
calc: $this * 100 / $1m_requests
54
units: %
55
every: 10s
64
- warn: ($1m_requests > 120) ? ($this > (($status >= $WARNING ) ? ( 1 ) : ( 2 )) ) : ( 0 )
65
- crit: ($1m_requests > 120) ? ($this > (($status == $CRITICAL) ? ( 2 ) : ( 5 )) ) : ( 0 )
56
+ warn: ($1m_requests > 120) ? ($this > (($status >= $WARNING ) ? ( 1 ) : ( 20 )) ) : ( 0 )
57
+ crit: ($1m_requests > 120) ? ($this > (($status == $CRITICAL) ? ( 20 ) : ( 30 )) ) : ( 0 )
58
delay: down 15m multiplier 1.5 max 1h
67
- info: the ratio of HTTP redirects (301, 303, 307, 308) over the last minute
59
+ info: the ratio of HTTP redirects (3xx except 304) over the last minute
60
to: webmaster
61
62
template: 1m_bad_requests
71
- on: web_log.response_codes
63
+ on: web_log.response_statuses
64
families: *
73
- lookup: sum -1m unaligned of 4xx
65
+ lookup: sum -1m unaligned of bad_requests
66
calc: $this * 100 / $1m_requests
67
units: %
68
every: 10s
77
- warn: ($1m_requests > 120) ? ($this > (($status >= $WARNING) ? ( 1 ) : ( 5 )) ) : ( 0 )
78
- crit: ($1m_requests > 120) ? ($this > (($status == $CRITICAL) ? ( 5 ) : ( 10 )) ) : ( 0 )
69
+ warn: ($1m_requests > 120) ? ($this > (($status >= $WARNING) ? ( 10 ) : ( 30 )) ) : ( 0 )
70
+ crit: ($1m_requests > 120) ? ($this > (($status == $CRITICAL) ? ( 30 ) : ( 50 )) ) : ( 0 )
71
delay: down 15m multiplier 1.5 max 1h
72
info: the ratio of HTTP bad requests (4xx) over the last minute
73
to: webmaster
74
75
template: 1m_internal_errors
84
- on: web_log.response_codes
76
+ on: web_log.response_statuses
77
families: *
86
- lookup: sum -1m unaligned of 5xx
78
+ lookup: sum -1m unaligned of server_errors
79
calc: $this * 100 / $1m_requests
80
units: %
81
every: 10s
@@ -132,38 +124,38 @@ families: *
124
# the following alarms trigger only when there are enough data.
125
# we assume there are enough data when:
126
#
135
-# $5m_2xx_last > 120
127
+# $5m_successful_old > 120
128
#
129
# i.e. when there were at least 120 requests during the 5 minutes starting
130
# at -10m and ending at -5m
131
140
-template: 5m_2xx_last
141
- on: web_log.response_codes
132
+template: 5m_successful_old
133
+ on: web_log.response_statuses
134
families: *
143
- lookup: average -5m at -5m unaligned of 2xx
144
- units: requests
135
+ lookup: average -5m at -5m unaligned of successful_requests
136
+ units: requests/s
137
every: 30s
146
- info: average successful HTTP requests over the last 5 minutes
138
+ info: average rate of successful HTTP requests over the last 5 minutes
139
148
-template: 5m_2xx_now
149
- on: web_log.response_codes
140
+template: 5m_successful
141
+ on: web_log.response_statuses
142
families: *
151
- lookup: average -5m unaligned of 2xx
152
- units: requests
143
+ lookup: average -5m unaligned of successful_requests
144
+ units: requests/s
145
every: 30s
146
info: average successful HTTP requests over the last 5 minutes
147
148
template: 5m_requests_ratio
149
on: web_log.response_codes
150
families: *
159
- calc: ($5m_2xx_last > 0)?($5m_2xx_now * 100 / $5m_2xx_last):(100)
151
+ calc: ($5m_successful_old > 0)?($5m_successful * 100 / $5m_successful_old):(100)
152
units: %
153
every: 30s
162
- warn: ($5m_2xx_last > 120) ? ($this > 200 OR $this < 50) : (0)
163
- crit: ($5m_2xx_last > 120) ? ($this > 400 OR $this < 25) : (0)
154
+ warn: ($5m_successful_old > 120) ? ($this > 200 OR $this < 50) : (0)
155
+ crit: ($5m_successful_old > 120) ? ($this > 400 OR $this < 25) : (0)
156
delay: down 15m multiplier 1.5 max 1h
157
options: no-clear-notification
166
- info: the percentage of web requests over the last 5 minutes, \
158
+ info: the percentage of successful web requests over the last 5 minutes, \
159
compared with the previous 5 minutes
160
to: webmaster
161