This commit is contained in:
@@ -2,81 +2,22 @@
|
||||
|
||||
PRIORITY - ALWAYS Lowest level stuff first, i.e. TODO at server, api route changes etc then back here in order lowest level first as affects the most stuff exponentially so best to do it early
|
||||
=-=-=-=-
|
||||
WIFI change 5g channel to 44 and 2g channel to 8
|
||||
WIFI change 5g channel to 52,56,60 and 2g channel to 8
|
||||
recheck before doing as it seems to vary, maybe someone else's is auto switching
|
||||
|
||||
https://www.datadoghq.com/blog/monitoring-101-collecting-data/
|
||||
|
||||
todo: OPS view metrics
|
||||
|
||||
Figure out what metrics to add to find bottlenecks and creeping issues and add them
|
||||
Am I running out of
|
||||
Disk space
|
||||
Memory
|
||||
CPU
|
||||
BANDWIDTH? (maybe not practical)
|
||||
avg response time / slowest routes?
|
||||
route metrics, slowest etc bottlenecks!
|
||||
exceptions per minute or something
|
||||
Disk usage
|
||||
Pie or donut chart with total available, used by attachments, used by backups
|
||||
Memory to longer time frame?
|
||||
reconsider all time frames
|
||||
|
||||
|
||||
RESEARCH:
|
||||
https://www.tpeczek.com/2017/06/feeding-server-timing-api-from-aspnet.html
|
||||
https://weblogs.asp.net/jeff/asp-net-core-middleware-to-measure-request-processing-time
|
||||
https://codereview.stackexchange.com/questions/235812/accurately-measure-asp-net-core-3-x-actions-execution-times-web-api-project
|
||||
https://stackoverflow.com/questions/48192756/how-to-get-actual-request-execution-time
|
||||
https://miniprofiler.com/
|
||||
https://github.com/MiniProfiler/dotnet
|
||||
DO stats for reference
|
||||
CPU USAGE
|
||||
LOAD AVERAGE (Processes waiting)
|
||||
MEMORY USAGE
|
||||
DISK USAGE
|
||||
DISK IO
|
||||
BANDWIDTH (PUBLIC)
|
||||
MINIPROFILER STUFF
|
||||
https://stackoverflow.com/questions/38249465/miniprofiler-how-do-i-profile-an-angularjs-webapi-app
|
||||
https://dotnetthoughts.net/using-miniprofiler-in-aspnetcore-webapi/
|
||||
https://stackoverflow.com/questions/53321634/unable-to-print-query-using-miniprofiler
|
||||
https://github.com/MiniProfiler/dotnet/issues/458
|
||||
https://community.miniprofiler.com/t/can-i-use-mini-profiler-for-asp-net-web-api-and-have-results-still-seen-on-url/365/2
|
||||
https://miniprofiler.com/dotnet/HowTo/ProfileCode
|
||||
https://miniprofiler.com/dotnet/AspDotNetCore
|
||||
|
||||
PROFILE VIEWING ROUTES
|
||||
http://localhost:7575/profiler/results-index
|
||||
http://localhost:7575/profiler/results-list
|
||||
http://localhost:7575/profiler/results
|
||||
|
||||
{http://localhost:7575/profiler/results-index?t=OquBJHsnjgWTyD5NQvAhTYcMWMeMVFaVjiQ7CBFwI}
|
||||
|
||||
todo: profiler
|
||||
Try out the hella big debug option, if it looks useful then
|
||||
Add a boot switch to turn on the hella big debug option or do it from the turn on and off route as well?
|
||||
|
||||
todo: Profiling
|
||||
Go over profiling and unit tests, see where there are obvious issues and bottlenecks now that I have the data
|
||||
Block licensing related queries and view and auth from profile
|
||||
Metrics view:
|
||||
Update: not sure I need this now, but maybe down the road might want to profile some particular thing outside of api routes?
|
||||
find a way to get the raw data so it can be fed back to the metrics route and limited to what I want to allow to show
|
||||
maybe store just the critical or slowest routes with some db info into the database on schedule from the in memory cache?
|
||||
https://github.com/MiniProfiler/dotnet/tree/master/src/MiniProfiler.Shared/ui
|
||||
|
||||
Full profiling view:
|
||||
|
||||
Add a filter to filter out anything auth related so no one can see that, i.e. filter out Authenticate routes where JWT is visible in query
|
||||
Add a link in UI for profiler, maybe host inside a frame or something
|
||||
(actually this might help with the auth dl token as well)
|
||||
Try out the hella big debug option, if it looks useful then
|
||||
Add a boot switch to turn on the hella big debug option or do it from the turn on and off route as well?
|
||||
|
||||
Our metrics not route related
|
||||
Can add our own metrics back into the profiling now that it's there with miniprofiler so if anything I want to track then can do that
|
||||
rockfish license fetching timer?
|
||||
maybe keep this in back pocket for now, not sure there is anything I would want to use it for
|
||||
|
||||
|
||||
|
||||
todo: Metrics, determine what else is needed for the basic graphing metrics, don't push it but whatever would be useful
|
||||
size of database as seen by postgres would be useful
|
||||
Look at pgadmin stats and see what is useful there
|
||||
|
||||
@@ -102,6 +102,7 @@ export default {
|
||||
"ServerJobs",
|
||||
"ServerLog",
|
||||
"ServerMetrics",
|
||||
"ServerProfiler",
|
||||
"NotificationSettings",
|
||||
"HelpAboutAyaNova",
|
||||
"MenuHelp",
|
||||
|
||||
Reference in New Issue
Block a user