add alarm for TCP/AttemptsFail spikes
Costa Tsaousis (ktsaou) committed
Jan 6, 2017 at 21:59 UTC
280dc08fd612fe8ec14143d577f83fb81a9bb4c5
2 files changed
+24
-3
conf.d/health.d/tcp_resets.conf
+23
-3
@@ -12,21 +12,41 @@
12
to: sysadmin
13
14
# -----------------------------------------------------------------------------
15
+# tcp resets this host sends
16
16
- alarm: 1m_ipv4_tcp_resets
17
+ alarm: 1m_ipv4_tcp_resets_sent
18
on: ipv4.tcphandshake
19
lookup: average -1m at -10s unaligned absolute of OutRsts
20
units: tcp resets/s
21
every: 10s
22
info: average TCP RESETS this host is sending, over the last minute
23
23
- alarm: 10s_ipv4_tcp_resets
24
+ alarm: 10s_ipv4_tcp_resets_sent
25
on: ipv4.tcphandshake
26
lookup: average -10s unaligned absolute of OutRsts
27
units: tcp resets/s
28
every: 10s
28
- warn: $this > ((($1m_ipv4_tcp_resets < 5)?(5):($1m_ipv4_tcp_resets)) * (($status >= $WARNING) ? (1) : (4)))
29
+ warn: $this > ((($1m_ipv4_tcp_resets_sent < 5)?(5):($1m_ipv4_tcp_resets_sent)) * (($status >= $WARNING) ? (1) : (4)))
30
delay: up 0 down 60m multiplier 1.2 max 2h
31
info: average TCP RESETS this host is sending, over the last 10 seconds (this can be an indication that a port scan is made, or that a service running on this host has crashed)
32
to: silent
33
34
+# -----------------------------------------------------------------------------
35
+# tcp resets this host receives
36
+
37
+ alarm: 1m_ipv4_tcp_resets_received
38
+ on: ipv4.tcphandshake
39
+ lookup: average -1m at -10s unaligned absolute of AttemptFails
40
+ units: tcp resets/s
41
+ every: 10s
42
+ info: average TCP RESETS this host is sending, over the last minute
43
+
44
+ alarm: 10s_ipv4_tcp_resets_received
45
+ on: ipv4.tcphandshake
46
+ lookup: average -10s unaligned absolute of AttemptFails
47
+ units: tcp resets/s
48
+ every: 10s
49
+ warn: $this > ((($1m_ipv4_tcp_resets_received < 5)?(5):($1m_ipv4_tcp_resets_received)) * (($status >= $WARNING) ? (1) : (4)))
50
+ delay: up 0 down 60m multiplier 1.2 max 2h
51
+ info: average TCP RESETS this host is sending, over the last 10 seconds (this can be an indication that a port scan is made, or that a service running on this host has crashed)
52
+ to: silent
configs.signatures
+1
@@ -278,6 +278,7 @@ declare -A configs_signatures=(
278
['c9b792755de59d842ba95f8c315d94c8']='health.d/swap.conf'
279
['ca026d7c779f0a7cb7787713c5be5c47']='charts.d.conf'
280
['ca08a9b18d38ae0a0f5081a7cdc96863']='health.d/swap.conf'
281
+ ['ca0eb92bdd3de67582ea6db37462895f']='health.d/tcp_resets.conf'
282
['ca249db7a0637d55abb938d969f9b486']='python.d/postfix.conf'
283
['cb178b15427274d7def5b14bc4c09441']='health.d/net.conf'
284
['cb60badf376d246ad8ec9d3f524db430']='health.d/disks.conf'