added bcache alarms
Costa Tsaousis (ktsaou) committed
May 29, 2018 at 01:52 UTC
ea1325f79d6e513f65c83204e114ceb2bc011073
3 files changed
+24
-1
conf.d/Makefile.am
+1
@@ -84,6 +84,7 @@ healthconfigdir=$(configdir)/health.d
84
dist_healthconfig_DATA = \
85
health.d/apache.conf \
86
health.d/backend.conf \
87
+ health.d/bcache.conf \
88
health.d/beanstalkd.conf \
89
health.d/bind_rndc.conf \
90
health.d/btrfs.conf \
conf.d/health.d/bcache.conf
new
+22
@@ -0,0 +1,22 @@
1
+
2
+template: bcache_cache_read_races
3
+ on: disk.bcache_cache_read_races
4
+ lookup: sum -10m unaligned absolute
5
+ units: races
6
+ every: 1m
7
+ warn: $this > 0
8
+ crit: $this > ( ($status >= $CRITICAL) ? (0) : (10) )
9
+ delay: down 1h multiplier 1.5 max 2h
10
+ info: the number of times bcache failed to read from the cache during the last 10 mins (this usually means your SSD cache is failing)
11
+ to: sysadmin
12
+
13
+template: bcache_cache_dirty
14
+ on: disk.bcache_cache_alloc
15
+ calc: $dirty + $metadata + $undefined
16
+ units: %
17
+ every: 1m
18
+ warn: $this > ( ($status >= $WARNING ) ? ( 70 ) : ( 90 ) )
19
+ crit: $this > ( ($status >= $CRITICAL) ? ( 90 ) : ( 95 ) )
20
+ delay: up 1m down 1h multiplier 1.5 max 2h
21
+ info: the percentage of cache space used for dirty and metadata (this usually means your SSD cache is too small)
22
+ to: sysadmin
src/proc_diskstats.c
+1
-1
@@ -1392,7 +1392,7 @@ int do_proc_diskstats(int update_every, usec_t dt) {
1392
, d->device
1393
, d->disk
1394
, family
1395
- , "disk.disk_bcache_cache_read_races"
1395
+ , "disk.bcache_cache_read_races"
1396
, "BCache Cache Read Races"
1397
, "operations/s"
1398
, "proc"