workflow create batch issues
hi,
I sent a batch of 5,000 workflow executions, and for some reason, many of them are stuck in a "QUEUED" state. No failures, no errors --just hanging there in that state
20 Replies
Can you provide more details? What is your account ID? What is the overall state of all of the Workflow instances?
I believe the account id is this:
998ad664b89c2f9efa99a3b294eadd9d
Uploading: Workflow A has not been executed yet.
Pending: Workflow A executed successfully and created a new instance for Workflow B.
Ready: Both Workflow A and Workflow B executed successfully.
Failed: One of the steps failed, likely due to other reasons—I still need to investigate.
Would it be better if there were a queue that managed the creation of workflow instances?
I ran an analysis on the instances that should have been created/executed, and here are some insights:
about half of them were not created -- probably because I exceeded the rate limit
however the ones that are queued up I don't know what's happening
hum from our side seems like stuff is running right now - as a suggestion I would use the
createBatch API , if possible, to reduce the number of calls and because of that we can better keep up with load for your accountsweet -- what happens with old ones that are still queued but not running?
all queued ones are from yesterday
they should eventually run - I found some things that we can improve on our control-plane to better suit your load profile 😅
Seems like your load profile is triggering a high amount of instances at the same time and then waiting a good while until the next batch?
sweet
I've adjusted a few things on my end to start using the
createBatch with 100 items per batch. Just give me a few and will let you know how it goes.
Also I'm curious about a few things, is the concurrent execution limit per account or per workflow?per account atm
mh.. look at this
sending 100 payloads with
createBatchyeah I saw it too, sending in a fix in a few - sorry about that 😅
all good
a fix is rolled out for this, let me know if it still breaks or something else is wrong - thanks :MeowHeartCloudflare:
sweet
thanks a ton
will check this out rn
the batch issue has been fixed, thank you
You can desconsider the number of queued ups, I'm cleaning this up
but as you can see the number or running instances slowly scale up
it has started with 1,000 and after ~20min it's been increasing up to 2,000
I’d like to understand what the behavior or rule is for increasing the number of running executions
I assume this happens automatically, but I couldn’t find more details on how it works
sweet -- after some time it has increased
