Lance Edgar
353abcc172
Make sure we use zero as fallback/default timeout values
2019-04-09 12:43:00 -05:00
Lance Edgar
cf27af81d4
Modify tempmon server logic to take "unfair" time windows into account
...
when a client or probe first are (re-)enabled, we can't expect to have readings
within the time window we'd normally be checking. previously we'd get false
alarms about "probe error status" etc. when this happened; hopefully no longer!
2019-01-25 19:49:46 -06:00
Lance Edgar
f31a0c4c22
Convert enabled for Client, Probe to use datetime instead of boolean
...
value is null if disabled, else timestamp of when it was last enabled
2019-01-25 19:33:49 -06:00
Lance Edgar
7212b07504
Fix bug when sending certain emails while checking probe readings
...
use common method to add more context for email template
2018-10-25 09:00:17 -05:00
Lance Edgar
30f0fe0a84
Add "default" probe timeout logic for server readings check
...
this way we don't have to set those timeouts on every single probe
2018-10-23 10:25:08 -05:00
Lance Edgar
19553edda6
Add per-status timeouts and tracking for probe status
...
i.e. this lets us keep track of when a probe becomes "high temp" and then later
if it becomes "critical high temp" we can still know how long it's been high
2018-10-19 14:58:30 -05:00
Lance Edgar
5830c7bd15
Fix logic bug when checking readings for client
...
yikes! first probe whose readings were checked okay, was causing other probes
to just be skipped
2018-10-17 19:26:48 -05:00
Lance Edgar
3aa4185de9
Leverage common "now" value when sending emails from server
...
i.e. don't generate new "now" when sending an email, just use the "now" we
established when starting to check the readings
2018-10-17 19:18:34 -05:00
Lance Edgar
71db57b2e0
Add probe URL to email template context
2018-10-17 18:08:54 -05:00
Lance Edgar
5bf69a0c21
Add "recent readings" to email template context
2018-10-08 00:52:16 -05:00
Lance Edgar
2f7fa3430a
Make server more tolerant of database restart
...
note that a retry is *not* attempted within a given "check readings" run.
rather, server will consider that full readings check to have failed, if any
part of it fails.
but then we keep track of type/amount of some (database connection) failures,
and will suppress logging the full error for first 3 attempts. in practice
this lets us recover from simple database restarts, and if database becomes
truly unavailable we'll hear about it shortly.
any other type of error is immediately logged on first failure.
2018-10-07 18:16:18 -05:00
Lance Edgar
b4fa6a17c5
Add "status since" to template context for email alerts
2018-10-06 20:17:46 -05:00
Lance Edgar
76e40063ee
Tweak server logic for checking client readings
...
do not check readings for "archived" clients. do not consider the client
"offline" unless it has *no* current probe readings. previously we were
assuming offline if any probe readings were missing, even if some were found.
2018-10-06 18:09:02 -05:00
Lance Edgar
018a9dcb08
Don't let server mark client as offline until readings fail 3 times in a row
...
previously we were only letting this fail once, if that
2018-09-28 11:57:27 -05:00
Lance Edgar
5df3379995
Send email alert when tempmon server marks a client as offline
...
courtesy of Cole Chaney <cole@mamajeansmarket.com>
2018-02-07 17:48:07 -06:00
Lance Edgar
8a1551e0f5
Send first alert "immediately" if critical temp status
...
i.e. only wait for "first email" delay if *not* critical
courtesy of Cole Chaney <cole@mamajeansmarket.com>
2018-02-07 17:47:48 -06:00
Lance Edgar
90e431c617
Switch license to GPL v3 (no longer Affero)
...
refs #2
2017-07-06 23:38:50 -05:00
Lance Edgar
396b3739f3
Fix bug when marking client as offline from server loop
...
..i think?
2017-06-01 17:37:27 -05:00
Lance Edgar
c4b371cedd
Refactor main server loop a bit, to add basic retry w/ error logging
...
hopefully this lets us get past a simple Postgres restart..
2017-06-01 17:16:31 -05:00
Lance Edgar
4e11748b45
Add configurable delay per client; improve try/catch
2017-02-09 17:59:06 -06:00
Lance Edgar
e6f3448acc
Hopefully fix alert logic when status becomes good
...
No need to send alert for good status, if previous status never actually
triggered an alert.
2016-12-11 10:46:53 -06:00
Lance Edgar
de1543018a
DRY up some server alert logic a bit
2016-12-10 23:04:42 -06:00
Lance Edgar
646f09a413
Add mail templates, plus initial status alert delay for probes
...
i.e. let the temp get "proper bad" (for some minutes) before alerting
2016-12-10 12:40:50 -06:00
Lance Edgar
fc7ec58303
Fix import bug in server daemon
2016-12-05 20:59:05 -06:00
Lance Edgar
7550e63940
Initial commit
2016-12-05 19:06:34 -06:00
