6b82684d5a
adapters: handle 204's on HttpAdapter
...
- adapters: add debug info on ping errors
2018-10-15 19:34:54 -03:00
37724f2d21
elstat.alerts: use ints instead of bool queries
2018-09-26 03:58:53 -03:00
867d7d1e69
elstat.adapters: only log on errors
...
always logging was giving quite the disk usage on prod.
2018-09-25 18:22:11 -03:00
9f5d42c89c
add downtime calculation on alerts
...
- add timestamp on alert embeds
- make Alert.post receive the worker instead of service info
2018-09-25 18:16:22 -03:00
227087284b
blueprints.api: use downtime in milliseconds, not seconds
...
This was giving way too optimistic availability rates.
2018-09-05 21:41:33 -03:00
93630f64ef
adapters: change "error not found" to "packet lost"
2018-09-05 21:23:58 -03:00
5428dcee3c
change uptimes to daily, weekly and monthly
...
- /api/current_status renamed to /api/quick
- add week_uptime and month_uptime to responses
2018-09-05 21:17:26 -03:00
e14720a8c8
remove print debug
2018-09-03 22:40:47 -03:00
3437bd2b72
Add better error messages
...
- Add status.json with HTTP status code table
- Add error fetching on PingAdapter
- Give warning on no alerts set
2018-09-03 22:39:41 -03:00
7a2f9a95c7
add alerts
...
- worker: change Adapter._construct to receive an extra arg
- worker: hardcode a bad result and write it off on workloop fail
2018-08-30 22:22:02 -03:00
3dd37c0bf1
small logging changes
2018-08-08 19:37:57 -03:00
d36d3d7449
blueprints.api: quickfix
2018-08-08 17:37:36 -03:00
86c99c8fee
blueprints.api: add calc_uptime, uptime field on api
2018-08-08 17:30:51 -03:00
93de3dfda0
worker: remove unecessary log call
2018-08-08 16:55:30 -03:00
6c94b50ba0
incidents: fix bug
2018-07-19 18:42:34 -03:00
slice
a44574249a
backend: INC_UPDATE -> INCIDENT_UPDATE
2018-07-18 15:06:31 -07:00
slice
d842ebfccb
give c field when publishing incident
2018-07-18 12:47:07 -07:00
slice
cea693102f
PUT /api/incidents -> POST /api/incidents
2018-07-18 12:46:58 -07:00
slice
11df05b998
unnest /api/incidents/current
2018-07-18 12:40:39 -07:00
slice
194129a5f2
/api/incident -> /api/incidents
...
for consistency
2018-07-18 12:39:43 -07:00
4407fbd6df
docs/streaming: add incident op codes and payloads
2018-07-18 05:01:32 -03:00
9c2dbeaf59
incidents: add rest of basic API
...
- change all ISO timestamp strings to unix timestamps
- add patch_incident and new_stage
- streaming: remove INC_UPDATE_STAGE
- manager: change primary key on instance_stages table
This requires a table change across all elstat instances, since
we changed the primary key on instance_stages.
2018-07-18 04:28:22 -03:00
dcadfd4c76
blueprints.incidents: add api base
...
- add getting a single incident, getting multiple (with pages), and
creating an incident
- blueprints.streaming: add incident related OP codes
- consts: add IncidentType
- manager: add incidents channel, add publish_incident
- elstat: add snowflake.py
2018-07-17 22:59:11 -03:00
e3a3e1b77b
worker: publish new status upon worker crash
2018-07-14 22:30:29 -03:00
83b2d88311
change adapter response type to dict
...
this simplifies a lot of code since tuples are immutable.
- adapters: add Adapter._construct
- worker: set status state to False on worker crash
2018-07-14 22:27:06 -03:00
ecf6234cfe
elstat.adapters: add latency to db spec
...
Since this is a change to the table schema for all services using the
ping adapter, you will have to delete the tables relating to them.
- elstat.blueprints: add incidents blueprint (WIP)
- elstat.blueprints: add decorators.py
- elstat.blueprints: add errors.py
- elstat.manager: add incidents & incident_stages tables
- run: add handler for ApiError
2018-07-13 23:20:09 -03:00
31a0f2b989
Add streaming API documentation
...
- streaming: add sub/unsub OP codes, changing from the old behavior of
just giving a channels array.
- streaming: properly check payloads against their data and send
proper errors.
- manager: add unsubscribe()
2018-07-11 18:23:47 -03:00
3fa27f5de6
move blueprints inside elstat folder
...
- add OP.DATA
- manager: add _websockets map
- manager: add basic publish logic
- worker: add timestamps to published messages
2018-07-11 17:00:39 -03:00
3ec64d5bdd
worker: add basic _dispatch_work logic
2018-07-11 03:42:31 -03:00
7e6d9d5402
add channel creation
...
dynamically create channels by adapter's specs.
- add cleanup tasks for ServiceManager and ServiceWorker
2018-07-11 02:48:00 -03:00
053762f69f
blueprints: add streaming blueprint
...
clients can connect and subscribe to different channels
and they will get messages from those channels (basic pub/sub)
this is useful in the case of a client wanting issue updates
in real time, which is better than the client having to poll
/api/status every now and then.
- manager: add ServiceManager.subscribe and .unsub_all
2018-07-10 20:03:07 -03:00
d92c09772e
add public api
...
- generalize elixire adapter into http adapter
- fix ping adapter not sending a tuple
- add ServiceWorker.process_work to insert results into the database
2018-07-09 02:48:44 -03:00
359c62efd6
Add proper adapter implementations
...
also add table making for the db
2018-07-08 00:38:14 -03:00
9ed1ec60cf
add ServiceManager & ServiceWorker
2018-07-07 20:51:43 -03:00
Luna Mendes
Luna Mendes
slice