timer job not triggering and activity functions are running out of resources.

[ghost]

Scenario: I have a timer job which wakes up every 1 hour and checks for any orchestrations running. if yes wont trigger a new one else it would. So i decided to have 4 orchestrations running at any time ( say with ids 1,2,3,4), each of them handling a specific task. Now each of this orchestrator calls in an activity function which reads data from azure table storage, reads a batch of signals and spawns 5 parallel threads and makes network calls. This keeps happening 24*7(each storage has around 1 million signals per day).

Running in app service plan with 2 instances.
ensured there is single cloud table client injected via startup, and using a http client factory for http calls.
issues facing: Timer job not triggering as per schedule. Running out of resources and only few orchestrators are making progress.

help needed: is it the thread issue? am i spawning too many threads. But the threads graph always showed 60, and the connection count was also less than 80 all the time.