(#2200) Callback verification failed with the following errors: curl_errno = 56; curl_error = CONNECT tunnel failed, response 429; HTTP Status Code = 429; HTTP Message = Too Many Requests
When activation succeeds (usually on 2nd or 3rd attempt), incoming WhatsApp messages do NOT trigger workflow executions.
Investigation & Proof
I checked both n8n and Traefik logs to trace what happens.
n8n logs (docker logs n8n --tail 30):
No mention of 429, Meta, or any webhook-related errors
n8n never sees the failed requests — they don’t reach the application
Traefik logs during failed activation:
No incoming request from Meta appears at all. Only my own browser requests (129.0.76.*) and healthchecks.
Traefik logs during successful activation (2nd attempt):
Meta’s verification GET request appears and succeeds:
172.19.0.1 - GET /webhook/[REDACTED]?hub.mode=subscribe&hub.challenge=[REDACTED]&hub.verify_token=[REDACTED] HTTP/1.1 200 9 - 54ms
Traefik logs after sending 3 WhatsApp messages (with workflow active):
ZERO webhook POST requests from Meta. Only healthchecks and my own browser polling for executions:
129.0.76.* - GET /rest/executions?filter=... HTTP/2.0 304
129.0.76.* - GET /healthz HTTP/2.0 200
129.0.76.* - GET /rest/executions?filter=... HTTP/2.0 304
129.0.76.* - GET /healthz HTTP/2.0 200
...(repeats with no Meta POST requests anywhere)
No execution was recorded in n8n despite messages being delivered (double blue ticks on WhatsApp).
Conclusion
Meta’s webhook POST requests are being blocked before they reach Traefik. Since Traefik has no rate limiting, n8n has no rate limiting, the OS has no firewall, and the Hostinger managed firewall has 0 rules — the only remaining layer is Hostinger’s own network infrastructure (DDoS protection or proxy layer).
This is NOT an n8n bug. Another user confirmed the same setup works perfectly on n8n Cloud with the same workflow. Multiple users on Hostinger VPS are experiencing this exact issue.
What I’ve already tried
Migrated from SQLite to PostgreSQL 16
Set N8N_DIAGNOSTICS_ENABLED=false, N8N_METRICS=false
Set N8N_CONCURRENCY_PRODUCTION_LIMIT=-1
Added Traefik forwarding timeout configurations
Added forwardedHeaders.insecure to Traefik entrypoint
Verified Hostinger managed firewall has 0 rules
Confirmed no UFW or iptables blocking
Environment
n8n version: 2.7.5
Database: PostgreSQL 16
Running via: Docker Compose behind Traefik v2.10
Hosting: Hostinger VPS KVM 2, Ubuntu 24.04
OS: Alpine Linux (inside Docker)
Has anyone on Hostinger found a workaround? I’m contacting Hostinger support and considering Cloudflare as a proxy layer.
The missing Traefik logs confirm Hostinger’s edge firewall is blocking Meta’s traffic before it even reaches your VPS.
Try this:
• Switch to Cloudflare: Move your domain’s DNS management to Cloudflare.
• Enable Proxy Mode: Turn on the “Orange Cloud” for your n8n subdomain to mask your server IP.
• Retest: This presents a trusted Cloudflare IP to Hostinger, usually bypassing their aggressive DDoS filters.
I’m dealing with the exact same issue — Instagram webhooks failing with 429 errors on a Hostinger VPS. Like you, we found nothing on our side generating the 429 (no rate limiting, no firewall, no WAF). After days of debugging, your post confirmed it: it’s Hostinger’s infrastructure-level rate limiting blocking Meta’s webhook deliveries before they even reach our server.
I saw Angela from Hostinger support said a fix was being rolled out. Did it actually fix the problem on your end?
No, they have not fixed the issue. And they cannot confirm when it will be fixed and they are relying on you to check regularly to see if the problem is fixed.
I have very little faith in them currently, because a week has passed and still no progress.
My solution was to change to a different hosting provider, which solves the problem.
The other solution that may work, but I have not tried it, is putting Cloudflare in front of your domain with proxy mode (orange cloud) enabled. This masks your server IP behind Cloudflare’s trusted IPs, which bypasses the aggressive DDoS filtering - as suggested by @achamm.
I can recommend changing to either n8n Cloud or different hosting provider. Or try putting Cloudflare in front of your domain with proxy mode (orange cloud) enabled. This masks your server IP behind Cloudflare’s trusted IPs, which bypasses the aggressive DDoS filtering. I have not tried the latter solution, but theoretically it should work.
I’ve decided I’m moving to Elestio. I followed using Hostinger because I saw many Youtubers recommended. Had a long talk with Claude and it has not reported any issues with Elestio and WhatsApp triggers unlike Hostinger and yes Hostinger is not keeping us updated with the situation and it is truly frustrating.
It was fun working with them but we just have to move on I guess. N8N Cloud is still expensive for me, so I’ll stick to self-hosted for now.
I can confirm I’m experiencing the exact same issue with Hostinger VPS.
Same setup:
n8n behind Traefik
Valid HTTPS/SSL
No firewall rules configured
Ubuntu on Hostinger VPS
Same error:
(#2200) Callback verification failed with the following errors: curl_errno = 56; curl_error = CONNECT tunnel failed, response 429; HTTP Status Code = 429; HTTP Message = Too Many Requests
The frustrating part is that it works intermittently - sometimes the webhook verification succeeds (usually on 2nd or 3rd attempt), but even when it activates successfully, incoming WhatsApp messages don’t trigger the workflow executions consistently.
I’m planning to open a support ticket with Hostinger requesting they whitelist Meta’s IP ranges. @henryford did you already contact Hostinger support? If so, how did you frame the ticket? Any specific information or approach that worked better to get their attention on this?
Also, has anyone tried the Cloudflare proxy workaround yet? Considering implementing it while waiting for Hostinger’s response.
I can confirm I’m having the exact same issue. I thought it might be a Meta problem, but after seeing this post, I can confirm that I’m also using Hostinger. I’m going to try a different server to see if it happens again.
We first started noticing 429 errors last week. A few days later, we realized that some WhatsApp triggers were not being registered in real time. Eventually, they were triggered several hours later, which caused users to receive random messages very late at night.
Yes, we are using Hostinger. We tried isolating the WhatsApp trigger to ensure it was properly returning a 200 status to Meta, but the issue persisted.
After reading this and other related forum posts, we are unsure whether migrating to Cloudflare or n8n Cloud would be the best solution. The migration would take several hours to set everything up again, and we’re not confident it would guarantee optimal functionality.
Has anyone successfully migrated to another hosting provider and resolved similar issues? Would submitting another support ticket or email potentially help?
Thanks in advance for your assistance.
For context, we are operating in the Mexico / North America region. We’re also unsure whether changing server locations could improve the situation.
Hola, mismo problema. Yo si tengo otro servidor, así que ahí realice pruebas y todo funciono bien. Habla con soporte de Hostinger para ver si podemos poner mas presión.
It doesn’t have anything to do with server location it’s a serious Hostinger issue and I’ll recommend you move to n8n cloud which I know is expensive or try elestio.
