alert store skeleton
Alert database type changes
Implement alert store for host alerts. All alert store skeletons.
Fix class method access
Enable tracing
Implements simple queries for host alerts
Implement flow alert store
Fixes escaping of INSERT queries
Flow alerts database schema fixes
Adds escaping for alert JSON in flows and hosts
Implements queries .select() for alerts store
Adds limit and offset to perform paginated queries
Adds new REST getter for flow alerts
Name changes alert_severity to severity, alert_json to json
Fixes alert message not shown
Implement active monitoring alerts store
Implements sort of queries in the new dataabase
Changes alert_type to alert_id
Implement mac alerts store
Fixes flow alert messages
Implement system alerts store
Implement snmp alert store
Add missing items to the flow alerts
Add missing items to the host alerts
Add missing items to the mac alerts
Implements COUNT aplerts api
Add device_name to snmp alerts
add flow alerts templates
updated gitignore
fix for missing order field
add families defined in `alert_store_schema.sql`
Reworks and simplifies alert store subclasses
Implements REST API to fetch alert timeseries
Fixes date in flow REST api
Host alert json fix
Add get/host/alert/list.lua
Move alert/list.lua to alert/past/list.lua
Add alert/past/list.lua for all alert families
Add entity_id to system table to identify the alert type based on <alert_id, entity_id>
Add missing field
Implements facilities to query engaged alerts via REST
Handle both historical and past alerts in alert/list.lua
Fix count
Update params of select_historical
Implement method to add family-specific filters
Add alert/ts.lua for all alert families
Implements facilities and REST endpoints to delete alerts
Implements ordering of alert queries
Fix add_order_by group_by
Rest API tests update
Tests output update
Remove debug trace
Use alert_id instead of type. Add more flow alert info.
Update http lint
Format obsolete tlv version alerts. Add more host info.
Add row_id to list of alerts
Fix selection of engaged alerts
Add test for mac alerts (bcast domains)
Removes attempt to format alerts as flow alerts
Fixes interface selection for active monitoring
Update test output with rest changes
Add more fields to be ignores
Set alert count to 1 for the time being
add bar timeseries chart
add apexcharts
improvements on timeseries bar chart
registered chart callbakcs
working on alert page
fix for date format
Fixes acrive monitoring REST API
Fixes alignment of grouped alert data
Additional fix for alert histogram
remove useless if
formatting alerts page
fixes on flows alert stats table
rename local networks to device
format host pagie in alert_stats
add tag support for hosts and flow
implemented single delete action
add release modal
Fixes format of threshold cross interface alerts
Fixes wrong increase of dropped alerts
Implements exclusion list for invalid dns queries
Reworks exclusions lists for hosts and flows
Addresses #5212
Addresses #5113
Adds host alert keys in host callbacks definitions
Adds alert ids to flow callbacks
fix for not working button (#5215)
Fixes reported timeseries name
removed any additional button inside chart's toolbar (#5200)
Add tables for interfaces, networks, users to the schema. Skeleton alert_store classes
fixed broken range picker layout in firefox (#5199)
Alert insert fixes
Add more info to network alerts
Add rest endpoint for interface, network, user alerts
Fix endpoint selection in alerts_stats
Unifies columns between engaged and past alerts
Fixes Missing mandatory 'alert_granularity'
Minor fixes for missing alert_severity
Fixes arithmetic on a nil value (field 'last_seen')
Fixes get/system/alert/list.lua
use tstamp for column names (#5221)
Implements host alert formatter
Add alerts_store format_record_common
Use common format_record for am, system alerts
Use common format_record for all alerts
Fixes formatting of alerts of all types
Fixes nil in function 'hostinfo2label'
fixes on disable modal
add pages for network, user and interface endpoint (#5224)
Set alert_entity in all classes
Unifies influxdb alerts into system alerts
Addresses #5224
Unifies process alerts into system alerts
Addresses #5224
Cleanup unused periodicActivityEntity
Unifies category lists alerts into system alerts
Addresses #5224
Aligns new alert enums
Addresses #5224
Fixes alert page links
Fixes insertion of interface alerts
Implement filters for Host alerts
Fixes active monitoring alerts not triggering
Implement filters on flow alerts
Fixes for internal alerts timestamp and subtype
implements disable for the alerts
formatted alert disable label
Add address and device type to mac alert records
fix for delete alert toggle
Fix access to entity_val in alert_unexpected_new_device
add mac address and device type inside table
Fixes for new alert fields not handled
Fixes alert_definitions to handle new fields
Add ip/port to snmp alert records
Implements deletion of stored flow alerts
Add alert_name to all alert records via rest. Fix duration.
fixes for snmp tab
Implements delete of past host alerts
Add name to snmp alert records
fixes on system tab
Fix Date column
fixes link
Update menu
Fixes bad argument #3 to 'format' in snmp alerts
updated interface link
new alerts url for host (#5228)
Fixes sort of engaged alert
Minor cleanup
Fixes data returned for local network alerts
Fix duration for one shot. Note.
Fix duration override
Fixes interface selection for system alerts
Move host alert page
fixes for local network tab
Minor fix
Fix engaged host alerts
fixes on user tab
Fix alert_user_activity message
While playing around with the timeseries feature I got the following exception when trying to use Telegraf's `inputs.influxdb_listener` as proxy to a Prometheus database.
```
...opng/scripts/lua/modules/timeseries/drivers/influxdb.lua:1643: attempt to index a nil value (local 'current')
```
`current` is likely supposed to be checked for nil earlier together with `required`, but instead of `current` the line is checking `version`, which isn't used anymore afterwards though.
So this commit changes the check from `version == nil` to `current == nil`.
Since then I've discovered that proxying Prometheus via Telegraf's `influxdb_listener` doesn't work anyways since ntopng also tries to read from it, but that's a different story.
