You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We're on a fork of an earlier version of wobserver, but I thought I'd bring this issue up in the core repo as an issue.
We've been seeing consistent timeouts from this parallel_map call: https://github.com/shinyscorpion/wobserver/blob/master/lib/wobserver/util/application.ex#L97. We've also noticed that the BEAM load spikes pretty hard whenever we navigate to the wobserver page for one of our apps that has many processes (~100 Ranch acceptors under a single supervisor). I'm guessing this has to do with all of the core :erlang calls getting process info
Here's a stacktrace from our version:
14:40:38.116 [error] Task #PID<0.1046.0> started from #PID<0.1033.0> terminating
** (stop) exited in: Task.await(%Task{owner: #PID<0.1046.0>, pid: #PID<0.1056.0>, ref: #Reference<0.0.1.97714>}, 5000)
** (EXIT) time out
(elixir) lib/task.ex:416: Task.await/2
(elixir) lib/enum.ex:1229: Enum."-map/2-lists^map/1-0-"/2
(wobserver) lib/wobserver/util/application.ex:101: Wobserver.Util.Application.structure_pid/1
(elixir) lib/task/supervised.ex:85: Task.Supervised.do_apply/2
(elixir) lib/task/supervised.ex:36: Task.Supervised.reply/5
(stdlib) proc_lib.erl:247: :proc_lib.init_p_do_apply/3
Function: #Function<1.105204156/0 in Wobserver.Util.Helper.parallel_map/2>
Args: []
14:40:38.116 [error] Ranch protocol #PID<0.1024.0> (:cowboy_protocol) of listener TaskRunner.Endpoint.HTTP terminated
** (exit) exited in: Task.await(%Task{owner: #PID<0.1046.0>, pid: #PID<0.1056.0>, ref: #Reference<0.0.1.97714>}, 5000)
** (EXIT) time out
When this time out happens the entire wobserver page fails to load which has made it unusable in app without changes.
I'd suggest using a pool to limit concurrency or chunking the processes
The text was updated successfully, but these errors were encountered:
Actually on further investigation, this is probably failing because we have a long-running message handler in one of our processes. Will report back when we get more info
We're on a fork of an earlier version of wobserver, but I thought I'd bring this issue up in the core repo as an issue.
We've been seeing consistent timeouts from this
parallel_map
call: https://github.com/shinyscorpion/wobserver/blob/master/lib/wobserver/util/application.ex#L97. We've also noticed that the BEAM load spikes pretty hard whenever we navigate to the wobserver page for one of our apps that has many processes (~100 Ranch acceptors under a single supervisor). I'm guessing this has to do with all of the core:erlang
calls getting process infoHere's a stacktrace from our version:
When this time out happens the entire wobserver page fails to load which has made it unusable in app without changes.
I'd suggest using a pool to limit concurrency or chunking the processes
The text was updated successfully, but these errors were encountered: