So my app uses a ton of API calls, on all non-critical automations using make (integromat) I have multiple delays to ease off the 10 requests per second limit. I have all my knack related steps set to retry 3 times because knack 503s constantly.
Up until now, I have had no issues with these and would only have some retries but mostly due to 503s and they would go through after.
In the past couple of weeks, every knack step is taking 3 tries with 429 rate limit errors, and maybe it will go through after that. Because of this, I have hit both my knack and make limits because I am essentially doing three times the operations every time. I have added additional, and longer delays to non-critical automations and they are still getting 429s.
We doubled our daily limit yesterday because our app went down when we hit our limit we have never come close to and now we are close to hitting this new limit.
What changed? I have not edited these automations in over a year.
I don’t know the answer but just wanted to check that you’ve raised a support ticket .
I’ll raise this with @Kara (here and the partner Slack channel).
She is Knacks new Product Manager and may be able to find out more.
Hello @JBS
Thanks for the post, and for tagging me, @CarlHolmes.
Could you please open a ticket with the @KnackSupport team, they will be able to gather the requisite information for us to investigate and figure out what’s going on for you.
Thank you,
Kara
Thanks Kara, it looks like the triggers I had using increment fields reset on Make, so our automations were trying to process every record every time they ran instead of just the new ones.
Is there a reason on knack’s end this could have happened? Make has not been very helpful, it started on the 26th and neither the table or automation had been edited in a few weeks.
Hi @JBS,
Not that I can find myself - Knack did not release any updates on Dec.26. If you open a ticket with Support, and let them know the table, details, and URL, they will be able to look into it more.
Thank you,
Kara
Thanks Kara, I finally got an answer from Make and it was an issue with their system.
This wreaked havoc on my app, since I have a bunch of automation that look for inconsistent data so they were all running and making API calls and using up operations.
Just letting you know in case anyone else has reported problems, I switched all my triggers to use webhooks instead of the increment field so it will not happen again.