#Pro plan monthly hard limit hit — production blocked, need grace reset

1 messages · Page 1 of 1 (latest)

steady meadow
#

Hi team,

Our production project (proj_wdtmrsovrdfysaqbysqr, account [email protected]) just hit the Pro plan monthly hard limit (5000/5000, error HARD_LIMIT_EXCEEDED). Our customers are currently blocked from triggering snapshots.

What happened:

We have a long-running batch polling task (cloro-chatgpt-batch-sender) that polls a 3rd-party scraping provider (Cloro). Two things went wrong this month:

2026-05-21 — netnut-batch-sender went into a runaway state, burning roughly $241 of compute in a single day. Root cause: provider downstream issues caused infinite-style polling.
2026-05-23 (today) — Cloro's polling API has a 3.5h queue backlog right now, so 30+ of our polling runs sat idle waiting for responses, each holding compute for 1-5 hours. This rapidly consumed the remaining budget.
What we've already done:

Cancelled all 30 stuck runs to stop the bleeding.
Rewrote the cloro-chatgpt-batch-sender task to be fire-and-forget: it now sends the batch + writes pending taskIds to KV in ~5 seconds and exits (vs 5h previously).
Moved the actual polling to a Cloudflare Worker scheduled cron (out of Trigger.dev compute entirely).
This is already deployed.
So the runaway pattern is fixed and won't recur. The new task uses roughly 3600× less Trigger.dev compute per snapshot.

Ask:

  1. Could you do a one-time grace reset of our monthly limit so our customers can use the service today? We're 1 week from the natural reset, and we're effectively offline.
  2. We'd also appreciate guidance on the next step up from Pro (before jumping to Enterprise). Are there higher-tier Pro-level options, or is Enterprise the only path for ~2-3× more compute headroom?
    Happy to share the deployment showing the fix is in place, and to discuss our usage pattern so we can size the right plan going forward.

Thanks,
Maciej Chmurkowski
Senuto

hollow plinthBOT
#

Hi @steady meadow. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

#

Hi Maciej,

Thank you for the detailed write-up — it's clear you've already done the right things to fix the root cause, and the architectural change (fire-and-forget + Cloudflare Worker cron) sounds like exactly the right approach.

However, I need to be upfront: I'm a documentation assistant and do not have the ability to reset account limits, access billing systems, or make account-level changes. You'll need to reach the Trigger.dev team directly for both of your requests.

How to get help urgently

Please contact the Trigger.dev team through one of these channels:

Make sure to include your project ID (proj_wdtmrsovrdfysaqbysqr) and account email in your message, exactly as you've described here.

What the docs say about higher limits

For context from the documentation:

  • Extra concurrency above the Pro tier limit is purchasable directly via the dashboard (Concurrency page in the left sidebar). [Limits]
  • Schedules beyond the Pro tier can be added in bundles for $10/month per 1,000. [Limits]
  • For rate limit increases, the docs note: "You can request a higher rate limit from us if you're on a paid plan." [Limits]

The docs don't explicitly describe a self-service path for increasing the monthly compute spend cap beyond Pro — that would need to be negotiated directly with the team.

Tips to avoid recurrence (already partly done by you)

#

The docs recommend several patterns you've now implemented or could consider: [How to reduce spend]

  • Use waitpoints instead of polling — waits >5 seconds checkpoint the task so you don't pay compute while idle.
  • Set maxDuration on tasks to cap runaway compute. [maxDuration]
  • Create billing alerts (75%, 90%, 100%, and spike alerts at 10x, 20x, etc.) to catch runaway usage early. [How to reduce spend]

I hope the Trigger.dev team can get you unblocked quickly — good luck!

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:

#

@steady meadow Kapa.ai is still learning and improving, please let me know how I did by reacting below.

steady meadow
#

Good morning. Can you help with the limit problem please?

tawny willow
#

Can you link to a run with this problem? I took a quick look at your account and I can't see it.

I'm pretty sure we don't emit a HARD_LIMIT_EXCEEDED error anywhere. Are you sure this error is coming from us and not the API you're hitting?

steady meadow
#

You're right. Sorry, false alarm- confirmed it's our own user-level monthly cap (5000 prompts/user/month, configurable per user in our DB), not yours. Our error code accidentally collided naming-wise with what you'd expect from Trigger.dev's quota system. Cleaning up. Thanks for the quick reply!
What a shame I didn't chek it with my dev team.