Interarrival as throttle #186
Conversation
a5f279b to
42d65c5
Compare
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #186 +/- ##
==========================================
+ Coverage 75.83% 77.03% +1.20%
==========================================
Files 32 32
Lines 1200 1167 -33
==========================================
- Hits 910 899 -11
+ Misses 290 268 -22 ☔ View full report in Codecov by Sentry. |
c16c3d6 to
2ae0719
Compare
73f1e61 to
92ef315
Compare
This also very importantly removes timers and loops from the critical controller process.
58f94c0 to
416b25e
Compare
416b25e to
8886183
Compare
| incr_no_of_users(SupNum) when SupNum > 1 -> | ||
| -spec handle_up_user(non_neg_integer(), pid(), amoc_scenario:user_id()) -> any(). | ||
| handle_up_user(SupNum, Pid, Id) when SupNum > 1 -> | ||
| ets:insert(?TABLE, {Pid, Id}), |
There was a problem hiding this comment.
can we just put user_id into the process dictionary of the user process? and add an interface at amoc_user to get user_id by pid (must be an rpc call for remote processes since process_info interface works only for the local processes)
There was a problem hiding this comment.
I'm not sure to understand what does that give. This is called by the supervisor worker, not the user process 🤔
src/users/amoc_users_sup.erl
Outdated
| %% Supervisor | ||
|
|
||
| %% @private | ||
| -spec start_link() -> supervisor:startlink_ret(). | ||
| start_link() -> | ||
| Ret = supervisor:start_link({local, ?MODULE}, ?MODULE, no_args), | ||
| UserSups = supervisor:which_children(?MODULE), | ||
| IndexedSupsUnsorted = [ {Pid, N} || {{amoc_users_worker_sup, N}, Pid, _, _} <- UserSups], | ||
| IndexedSupsUnsorted = [ {Pid, N} || {{amoc_users_worker_sup, N}, Pid, _, _} <- UserSups, |
There was a problem hiding this comment.
could you remind me why it's unsorted?
There was a problem hiding this comment.
and please add a comment at indexes/0 that we start indexes from 2 to simplify user_count atomics management.
There was a problem hiding this comment.
They'll be returned in reversed order from where they were started, so if they were declared as 1..12, this will return 12..1, so maybe even a reverse would do. But just to be safe I thought of a keysort.
| Objects | ||
| end, | ||
| stop_children_assignments(Users, Force), | ||
| length(Users). | ||
|
|
||
| -spec get_all_children() -> [{pid(), amoc_scenario:user_id()}]. | ||
| get_all_children() -> |
There was a problem hiding this comment.
seems that this function is used for testing only, we can remove it and rework testing a bit.
There was a problem hiding this comment.
Not sure how without literally doing ets:tab2list from tests instead of here, but that'd make this one a leaky abstraction 🤔
No description provided.