We are exploring what we can and cannot achieve with Munchkin calls using the visitWebPage, clickLink and associateLead functions. A question came up during brainstorming, "Are Munchkin calls throttled?" A broad question and one without specifics, but I thought knowing a little more about how Munchkin behaves would be useful.
Is there an upper daily limit (I was once told 250,000)?
What number of calls within what timeframe would trigger a throttle if it exists?
🙂
Solved! Go to Solution.
Hi Alan,
The daily limit with Munchkin 1.0 (still the current version) is said to be 2 million "visits," which in context seems like it means "activities" (i.e. VisitWebPage, Click Link, and Associate Lead calls). In Munchkin 2.0 this is upped to 100 million of that same unit per instance per day. Getting on Munchkin 2.0 sooner than the bulk of users may be a bit of a chicken-egg situation for you, since you have to demonstrate load. Eventually, though, everyone will be on 2.0.
As for point-in-time throttling, I'm not aware of any published limit that is lossy, though I know there's a load limit beyond which the servers send back TCP resets instead of completing the HTTP response. Supposedly, this situation is still lossless, since only the request phase of the GET is necessary for servers to receive the data.
But it isn't only about how many activities can be logged over the course of a day, but whether those activities actually make it to the MLM back end in time to be observed and/or actioned, instead of being backlogged (though not technically "lost"). Munchkin 2.0 is supposed to resolve these concerns as well, but in the meantime I'd advise the strategic disabling of redundant activities. For example, if the target URL of a Click Link is itself running Munchkin (i.e. is not an offsite and/or binary asset), disable the click tracking on that link and log only the subsequent Visit Web Page. These measures free up calls to be used for more important actions, say logging video play/pause events, which aren't redundant.
Hi Alan,
The daily limit with Munchkin 1.0 (still the current version) is said to be 2 million "visits," which in context seems like it means "activities" (i.e. VisitWebPage, Click Link, and Associate Lead calls). In Munchkin 2.0 this is upped to 100 million of that same unit per instance per day. Getting on Munchkin 2.0 sooner than the bulk of users may be a bit of a chicken-egg situation for you, since you have to demonstrate load. Eventually, though, everyone will be on 2.0.
As for point-in-time throttling, I'm not aware of any published limit that is lossy, though I know there's a load limit beyond which the servers send back TCP resets instead of completing the HTTP response. Supposedly, this situation is still lossless, since only the request phase of the GET is necessary for servers to receive the data.
But it isn't only about how many activities can be logged over the course of a day, but whether those activities actually make it to the MLM back end in time to be observed and/or actioned, instead of being backlogged (though not technically "lost"). Munchkin 2.0 is supposed to resolve these concerns as well, but in the meantime I'd advise the strategic disabling of redundant activities. For example, if the target URL of a Click Link is itself running Munchkin (i.e. is not an offsite and/or binary asset), disable the click tracking on that link and log only the subsequent Visit Web Page. These measures free up calls to be used for more important actions, say logging video play/pause events, which aren't redundant.
Thanks for the thorough explanation Sanford! Very helpful.