iankelling.org
/
git
/
distro-setup
/ blobdiff
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
shortlog
|
log
|
commit
|
commitdiff
|
tree
raw
|
inline
| side by side
mostly fixes, some new scripts
[distro-setup]
/
filesystem
/
etc
/
prometheus
/
rules
/
iank.yml
diff --git
a/filesystem/etc/prometheus/rules/iank.yml
b/filesystem/etc/prometheus/rules/iank.yml
index 0ee3335fdbd76bef63547daca50fc9bc41d321e6..971e392fd63b55387089c47529060ebd91543d03 100644
(file)
--- a/
filesystem/etc/prometheus/rules/iank.yml
+++ b/
filesystem/etc/prometheus/rules/iank.yml
@@
-44,7
+44,10
@@
groups:
###### END MISC NOTES ######
###### END MISC NOTES ######
-
+# various queries only look at increases, so invert the up metric so we
+# can better query on down.
+ - record: down
+ expr: up == bool 0
# alerting on missing metrics:
# alerting on missing metrics:
@@
-112,15
+115,6
@@
groups:
annotations:
summary: '12 minutes down'
annotations:
summary: '12 minutes down'
- # 42 mins: enough for a 30 min queue run plus 12
- - alert: mailtest_check_vps
- expr: |-
- time() - mailtest_check_last_usec{job="tlsnode"} >= 60 * 42
- labels:
- severity: prod
- annotations:
- summary: '42 minutes down'
-
- alert: mailtest_check_mailhost
expr: |-
time() - max by (folder,from) (mailtest_check_last_usec{job="node"}) >= 60 * 12
- alert: mailtest_check_mailhost
expr: |-
time() - max by (folder,from) (mailtest_check_last_usec{job="node"}) >= 60 * 12
@@
-130,9
+124,9
@@
groups:
summary: '12 minutes down'
# 42 mins: enough for a 30 min queue run plus 12
summary: '12 minutes down'
# 42 mins: enough for a 30 min queue run plus 12
- - alert: mailtest_check_mailhost
+ - alert: mailtest_check_
gnu_
mailhost
expr: |-
expr: |-
- time() - max by (folder,from) (mailtest_check_last_usec{
job="node
"}) >= 60 * 42
+ time() - max by (folder,from) (mailtest_check_last_usec{
folder="/m/md/l/testignore", from="iank@gnu.org
"}) >= 60 * 42
labels:
severity: prod
annotations:
labels:
severity: prod
annotations:
@@
-175,11
+169,11
@@
groups:
# avg_over_time(node_systemd_unit_state{name="dynamicipupdate.service",state="active"}[1d]) < .95
- alert: up_resets
expr: |-
# avg_over_time(node_systemd_unit_state{name="dynamicipupdate.service",state="active"}[1d]) < .95
- alert: up_resets
expr: |-
- resets(up[
2d]) - changes(node_boot_time_seconds[2
d]) > 12
+ resets(up[
1d]) - changes(node_boot_time_seconds[1
d]) > 12
labels:
severity: warn
annotations:
labels:
severity: warn
annotations:
- summary: "Target has gone down {{ $value }} times in
2 days
, > 12"
+ summary: "Target has gone down {{ $value }} times in
1 day
, > 12"