Closed gsreynolds closed 1 year ago
Substantial performance improvements by @martindstone around how PD Live implements, handles and queues async requests as well as terminates running async requests in response to queries being changed.
https://github.com/giranm/pd-live-react/pull/453
switch back to pdParallelFetch for getIncidents in redux/incidents/sagas to improve performance and stop making too many requests
get rid of useless nested loop in pdParallelFetch (my bad)
use throttledPdAxiosRequest in pdParallelfetch
re-filter the list of incidents in refreshIncidents to stop making too many calls for notes and alerts
set more sensible Bottleneck defaults (200/min instead of 1900/min)
added API call rate limit (100-2000) as a user configurable setting
stop truncating the incidents list to max incidents limit (no need to do this when we rate limit appropriately)
change the query settings modal to warn that fetching will be slow instead of saying only X incidents will be retrieved
abort any in-flight notes and alerts requests when getting or refreshing incidents
show the number of queued and running API calls in the status beacon tooltip
Substantial performance improvements by @martindstone around how PD Live implements, handles and queues async requests as well as terminates running async requests in response to queries being changed.
https://github.com/giranm/pd-live-react/pull/453
Details of async request improvements
switch back to pdParallelFetch for getIncidents in redux/incidents/sagas to improve performance and stop making too many requests
get rid of useless nested loop in pdParallelFetch (my bad)
use throttledPdAxiosRequest in pdParallelfetch
re-filter the list of incidents in refreshIncidents to stop making too many calls for notes and alerts
set more sensible Bottleneck defaults (200/min instead of 1900/min)
added API call rate limit (100-2000) as a user configurable setting
stop truncating the incidents list to max incidents limit (no need to do this when we rate limit appropriately)
change the query settings modal to warn that fetching will be slow instead of saying only X incidents will be retrieved
abort any in-flight notes and alerts requests when getting or refreshing incidents
show the number of queued and running API calls in the status beacon tooltip