How to handle API calls in a row in Locust ? about locust

[chunji08]

What do you mean "the api call hangs until the time running out"?

In the middle my Locust running, if the response1 failure happened, and I had it "return", then till the end of this locust running, I don't see any extra response1 or response2 calls are made.

Thanks for the tips, I will give a try.

Thanks.

Jack

[cyberw]

If I was to have it executed with extra users, in some running session, the first API call hits the failure, and return back 400 code. In my current code, I was doing "return", which hopes the current session to be stopped, and a new one would be started. But I don't see it happens. What I have noticed is, all the API call hangs until the time running out.

What do you mean "the api call hangs until the time running out"?

I recommend running the test in a debugger or adding some print statements.

a) return is fine, should work pretty much the way you expect, unless you mean you want to run on_start again
b) dont throw StopUser, it is an internal thing (we should probably rename it _StopUser or something...)
c) not really. The one thing that I think you may be confused about is the fact that Users loop forever. Their session is never terminated and on_start only happens once. If you need fine grained control use a flag:

class MyUser(HttpUser):
    first_run = True

    @task
    def t(self):
        if self.first_run:
            self.first_run = False
            # Initialization code goes here      
        ...
        # if current call fails, needs to restart from scratch.  
        if response1.status_code != 200:
           self.first_run = True
           return

Another way to do it is just to call on start directly:

    # if current call fails, needs to restart from scratch.  
    if response1.status_code != 200:
       self.on_start()
       return