Hi everyone,

I created a new credential through the render dashboard. I’m trying to rotate to this new credential and delete the old one.

When using this new role, I get the error:

permission denied to cancel query Only roles with privileges of the role whose query is being canceled or with privileges of the “pg_signal_backend” role may cancel this query.

This does not happen with the original role. What is the proper and correct way of handling this? I want this new credential have the exact permission and privilege as the old one.

As we look to the future, we’re excited to bring our community together in a more dynamic space. We’ll be sunsetting our Discourse forum on 24th March 2026 so we encourage you to save any threads or solutions you’d like to hold onto before then. We can’t wait to continue the conversation and collaboration on Discord. Come say hello and join us at Render Developers . See you there!

Your account has been suspended for suspicious activity. If you believe this was a mistake, please contact support.

I sent an email to support and received the following:

Hi there,

Your account has been blocked due to one or more violations of our Acceptable Use Policy.

We’ve reviewed the services that were deployed and will not remove this block

I have no clue what or why my account was suspended and unfortunately, I haven’t gotten a response to my query. Obviously, I’m not intentionally trying to violate the policy, but it would be nice to know what the violation is in regards to and a way to retify it.

One service was spun down and the other was suspended by me, so neither should have been consuming any time. Earlier today (around 10:30 ish am), the usage was at 0.1/750 hours. I also checked the logs for both services, and they weren’t used at all except for a test I did at 1 AM. After that, both services spun down followed by one service being suspended.

Aside from that, they weren’t running, worked on, or pinged.

So now I’m now confused - do suspended services consume instance hours? Could the usage have come from the spun-down service?

Any clarity/explanation would be appreciated.

If you change from the Hobby workspace plan to the Pro Workspace plan, then your service will still be using a starter instance.

Workspace plan controls features you have access to, Service instance type controls compute resources available.

We have a hubby plan but our Node.js server is on a starter Instance Type. If we scale our plan from Hubby to Professional ($19/month), which instance type will our Node.js server be in? Will it be automatically upgraded to Standard Instance Type or Pro Instance Type?

I am writing to formally appeal the suspension of my account (associated with this email and GitHub: [ saab470 ]). My account was recently flagged for “suspicious activity”, and I believe this is an automated misunderstanding regarding my project, Atlantic Store (atlantic- store. onrender. com).

I want to clarify the following points for your manual review:

This is a personal portfolio project built to practice React, Socket. io, and Full-Stack development. It is not a live commercial business.

The cryptocurrency images and phrases like “return capital and profit” are purely simulated UI elements used to test order-processing logic. There are no real financial transactions occurring on this platform.

I noticed some broken image links in my dashboard. I am in the process of fixing these and updating the site metadata from the default “React App” to a proper portfolio title.

I am a student developer aiming to follow all Terms of Service. If any specific content (images or text) is triggering your security filters, I am more than happy to remove or replace it immediately to ensure compliance.

I put a lot of work into the coding logic of this store and would greatly appreciate it if a human team member could review my account and restore access.

Thank you for your time and for keeping the platform safe.

Best regards,

[Muhammad Bilal]
——————
I sent email but it says that blocked.
Your message to support@ render. com has been blocked. See technical details below for more information.
LEARN MORE
The response was:
Message rejected.

Also trying to attach screenshots to the post but it says I cant attach media

I just deleted the web service and recreated it in a different region, and the issue no longer occurs. The problem only persists when I deploy the web service in US East - Ohio.

My application is running normally in US East - Virginia. However, this is not ideal since the Supabase servers are located in Ohio.

Could someone please look into this issue?

A few days ago, an application I’m developing—which had been working normally—started showing an intermittent error when loading pages:

status: 500
message: Knex: Timeout acquiring a connection. The pool is probably full. Are you missing a .transacting(trx) call?

The error is inconsistent. Most of the time, the application shows this error, but sometimes it starts working again as it should, with good performance. However, at some point, it simply goes back to the same error… and stays that way for several minutes until a “miraculous” refresh brings the app back, and the cycle repeats.

Initially, I thought it was a database server issue (Supabase - OHIO), but I realized the problem only affects the communication between Render and Supabase. In my local environment, which connects to the same database, this error never happens!

I have an outdated version of the same application (which will be decommissioned soon) that also connects to Supabase, and no issues occur there.

What could be happening? Could you help me out?

Thanks,
Ivan.

.

And from live chat:

A database doesn’t have “previews” in the same way that, say, a web service does where the base service lists all of the previews of itself (whether by PR preview or preview environments) under /previews . So, it looks like a database in a preview environment is always listed even without the flag since it’s not technically a preview but is part of a preview environment.

• web
• jobs
• docker image
• database 1
• database 2

If I do render services I only see the 2 database listed, for my production app I see all 5 services.

The services are up and running for the PR preview and I only have 1 page of services in the CLI results. This means if I want to SSH into the web service to do something I must copy the SSH connection from the render dashboard, which does connect but gives me all sorts of weird connect failed output and sometimes just disconnects my session. Why aren’t they showing up in the CLI?

When trying to connect my github to create a new service, it redirects me to the Render app permission in Github. I give all perms and then when back and click the connect github again, it redirects me to the perms page again, it’s an infinite loop. What I have done:

1. Disconnected github account and reconnected it
2. Hard cache reset

Any ideas?

Thank you.

1. I have a 3rd party managed OpenSearch provider in use on production, for previews I would be fine to have OpenSearch spun up from a docker image. Is it possible to have preview only specific services.
2. The env variables are coming from an environment group, I’ll need some to be overridden/different for preview env, how can I do that? Ideally having a preview environment group would be best.

==> No open ports detected, continuing to scan…

==> Docs on specifying a port: https://render.com/docs/web-services#port-binding

==> Port scan timeout reached, no open ports detected. Bind your service to at least one port. If you don’t need to receive traffic on any port, create a background worker instead.

==> Docs on specifying a port: https://render.com/docs/web-services#port-binding

**==> Timed Out
**
i have already made a PORT environment variable and set it to 8080. i don’t have anything in my rust scripts that is related to http requests, so how would i add that

Tell me how to reach them and ask for refund.

Thanks.

My website is legitimate and does not impersonate the brand referenced in the domain in any way.
I replied to Render support right away but haven’t received a reply, and it’s been over 6 hours. I’m worried I got caught up in some kind of automated security check, but I can’t reach anyone about this.

This is certainly making me concerned about relying on Render as a platform moving forward or recommending it to anyone I know.

I’m running n8n self-hosted on Render, managing workflows via the **n8n MCP server** (Model Context Protocol for AI-assisted automation). The underlying issue is WAF blocking REST API calls, but the symptom first appeared as MCP tool failures.

**The Issue:**

When updating workflows that contain multiple Code nodes (3+) with JavaScript, requests return 403 Forbidden with a Cloudflare WAF block page. Smaller payloads with the same structure succeed.

**How it manifests:**

- **Via n8n MCP:** `n8n_update_full_workflow` or `n8n_create_workflow` fails with 403

- **Via REST API:** PUT `/api/v1/workflows/{id}` returns Cloudflare block page

- **Via n8n UI:** Works fine (same operations succeed)

**What I’ve Tried:**

- Smaller payloads (1-2 Code nodes): Works

- Manual updates via n8n UI: Works

- Incremental node addition via MCP: Sometimes blocked mid-sequence

**Environment:**

- n8n self-hosted on Render (Hobby plan)

- n8n MCP server for programmatic workflow management

- Also affects direct REST API calls

**Error:**

- HTTP 403 Forbidden

- Cloudflare WAF block page (not n8n API error)

- CF-RAY: `9c88dd08b97a0731`

**Questions:**

1. Anyone else using n8n MCP or REST API on Render hitting this?

2. Any known workarounds for programmatic workflow management?

3. Is there a way to request WAF tuning for specific endpoints?

-–

## How to Check if You’re Affected

1. Host n8n on Render

2. Use n8n MCP or REST API to create/update a workflow

3. Include 3+ Code nodes with `jsCode` parameters (~500+ chars each)

4. Total payload ~2.5KB+

5. Check if you get HTTP 403 with Cloudflare block page

**How to get your CF-RAY (trace ID):**

```bash

curl -i -X PUT “https://[your-service].onrender.com/api/v1/workflows/[id]” \

-H “Content-Type: application/json” \

-H “X-N8N-API-KEY: [key]” \

-d ‘{“name”:“test”,“nodes”:[…]}’ 2>&1 | grep -i “cf-ray”

```

Or check MCP error output - the CF-RAY may be in the response body.

**If you hit this issue**, please reply with:

- Your CF-RAY ID

- Your use case (n8n MCP, direct API, Claude Code, etc.)

More data points help Render prioritize a fix.

-–

**Support ticket submitted:** Via Render chatbox. no ticket number provided.

Both can’t find the image, but in the UI, the displayed docker image, when clicked, goes to the proper page on docker hub

At the top of the blueprint page it shows the correct docker image (and when clicked goes to the proper place), but can’t be found for the image pull

I can’t post a GitHub url, so here’s the render.yaml i’m using:

# Exported from Render on 2026-02-02T13:59:43Z
version: "1"
databases:
- name: fusionauth
  databaseName: fusionauth_db
  plan: free
  ipAllowList:
  - source: 0.0.0.0/0
    description: everywhere
  postgresMajorVersion: "18"

services:
- type: web
  name: fusionauth-app:latest
  runtime: image
  image:
    url: fusionauth/fusionauth-app:latest
  plan: free
  envVars:
  - key: DATABASE_URL
    fromDatabase:
      name: fusionauth
      property: connectionString
  - key: POSTGRES_PASSWORD
    fromDatabase:
      name: fusionauth
      property: password
  - key: POSTGRES_USER
    fromDatabase:
      name: fusionauth
      property: user
  - key: FUSIONAUTH_APP_RUNTIME_MODE
    value: development
  - key: DATABASE_USERNAME
    fromDatabase:
      name: fusionauth
      property: user

error: failed to solve: process “/bin/sh -c python manage.py collectstatic --noinput” did not complete successfully: exit code: 1

I’m struggling to get a Celery worker on Render to connect to my self-hosted RabbitMQ instance on an external VM via Tailscale. I’ve been debugging this for 3 days and keep hitting a wall with [Errno 111] Connection refused.

The Setup:

Service Type: Background Worker.

Networking: Using Tailscale in userspace-networking mode (since Render doesn’t support TUN).

The Goal: Connect the Celery worker on Render to RabbitMQ running on a VM

The Problem: Tailscale seems to initialize correctly, and I can see the Render node in my Tailscale dashboard. However, Celery fails to connect to the broker.

Error Log:
[2026-01-31 12:48:12,044: ERROR/MainProcess] consumer: Cannot connect to amqp://user:**@127.0.0.1:5672//: [Errno 111] Connection refused.

My start-worker.sh: I am using tailscale nc to bridge the remote RabbitMQ port to localhost, but Celery still can’t “see” it:

./tailscaled --tun=userspace-networking --socket=./tailscaled.sock --state=mem: --no-logs-no-support &
sleep 15

./tailscale --socket=./tailscaled.sock up --authkey=$TAILSCALE_AUTHKEY --hostname=cove-worker --accept-dns=false

./tailscale --socket=./tailscaled.sock nc ace-01 5672 &

echo “Bridge established: localhost:5672 → ace-01:5672”
sleep 5

exec celery -A my_project --broker=“pyamqp://user:[email protected]:5672//” worker -Q gateway_queue --loglevel=info --pool=solo

**What I’ve tried:
**
* Using the Tailscale IP (100.x.x.x) directly in the broker URL.
* Using a SOCKS5 proxy (socks5h://localhost:1055) via CELERY_BROKER_TRANSPORT_OPTIONS.
* (Verified that RabbitMQ on the external VM is listening on all interfaces (0.0.0.0) and the firewall allows port 5672.
* Switching to a unique hostname for each deploy to avoid 404: node not found errors.

And yes I made sure to install Tailscale during build phase

My Question: Does Render’s Background Worker environment restrict internal TCP bridging/loopback traffic? Is there a specific way to ensure the Python amqp library respects the userspace proxy or the tailscale nc bridge?

i contacted render support . i cannot login render.com dashboard due to no access to 2fa codes on my mobile device since it was factory reset. now i cant access my website itself. getting 502 error now. i have been using the service for a long time. now i am getting these issues. please help me resolve the issue

thanks

Thanks

I’m running a backend service on Render that uses Inngest for background jobs, and over the last ~month I’ve started seeing intermittent failures that look like this:

Error performing request to SDK URL:
Post "https://{service}.onrender.com/inngest/register?fnId={pdf-parse-function}&stepId={step}":
dial tcp [64:ff9b::d818:39fb]:443: i/o timeout

A few important details / observations:

• This is intermittent, not consistent

  • Same service, same endpoint

  • Some Inngest runs succeed, others fail minutes apart

• Failures appear during the Inngest SDK registration / handshake, before function logic runs

• The IP (64:ff9b::/96) looks like a NAT64 / IPv6-mapped address

• No correlation with load, concurrency, or deploys

• Increasing timeouts / retries inside the function does not help (this happens before execution)

What’s confusing is:

• This setup worked reliably before

• Nothing significant changed on our side when this started

Any Ideas how I can go ahead debugging this.

I had a fastify app running happily on render.
However, since I migrated to express (because of the microsoft bot framework)

@teams-vote/[email protected] start /opt/render/project/src/src/teams-vote-api
Menu
node dist/server.js
API running on { address: ‘0.0.0.0’, family: ‘IPv4’, port: 10000 }

If I open my health page it does nothing and the deploy logs just hang at this point.
I don’t know what to do next.

Docker recently made their Docker Hardened Images free for use. This is great, because it allows many people to benefit from greater security in their applications.

Unfortunately, they’re gated behind a new registry they set up that’s separate from docker hub, meaning they cannot be pulled by Render, which only supports a few preconfigured registries.

Is there any plan to add support for custom docker registries? Or docker’s hardened images?

We’re currently evaluating where we’re going to deploy our HIPAA app, with the main contenders being Railway and Render, and this is a bummer / a point in favor of Railway.

Thank you!

-Ted

I’ve been having a problem only when I run a nuxt app using a docker image.

Sometimes work, and sometimes it doesn’t.

Sometimes it doesn’t find the resources and after F5, it does… it’s really strange the behavior.

Is there a specific environment variable I need to add to make it work?

Thanks

**
{**“message”: “Not Found - /”,“stack”: null}

• This is a Go API service connecting to an external postgres database.
• Logs don’t show any errors at the time of failure. There’s also no indication the app restarted after, it just kept on working
• Memory at 5% and CPU at 0% at time of failure

Appreciate any guidance on troubleshooting what seems to be issue that’s not appearing anywhere. Thanks.

Would appreciate help from others who may have run into the same issue too

I want to deploy my Python web app in Render Cloud.

This web app will use Flask to listen for webhook from Stripe.

The questions I have is: Will the Render Cloud server sleeps if there are very few transactions to my web app?

Reason: This is because I am worried that my web app cannot receive the webhook from Stripe.

Thank you.

John