Skip to content
High Latency and API Timeouts from Bangalore, India – Singapore Region
lazizia
OP

2 months ago

Hi Railway Support Team,

I’m frequently experiencing very slow response times when accessing the Railway platform and my application from Bangalore, India. My application and database are hosted in the Singapore region.

Everything was working fine earlier, but recently I’ve started experiencing significant slowness, and sometimes API requests are taking too long or timing out.

Could you please investigate this issue and fix the underlying connectivity/performance problem? It seems to be an issue that has started recently, as it was working normally before.

Thank you.

Solved

8 Replies

Railway
BOT

2 months ago

This thread has been opened as a bounty so the community can help solve it.

Status changed to Open Railway • about 2 months ago


Anonymous

2 months ago

Same thing from India this morning, against the Singapore region, in bursts. It is not your app being slow.

Run these two lines while it feels slow:

curl -s -o /dev/null -w "app tls=%{time_appconnect}s total=%{time_total}s\n" 'https://YOUR_APP_DOMAIN/health'

curl -s -o /dev/null -w "google tls=%{time_appconnect}s total=%{time_total}s\n" 'https://www.google.com'

Then open your service HTTP logs for that same minute.

What we saw, on a deployment that was otherwise idle:

  • Google from the same laptop: about 0.2s
  • railway.com and our Singapore app, from that laptop: often 10–20s, most of it in the TLS handshake
  • The API container’s own log for the same request: about 50ms
  • A static file the server logged at 1ms still took many seconds on the client

So the handler had already finished. The missing seconds were between Bangalore and the container, and they came and went. Rough windows today (IST): about 10:00–10:30, 10:50–11:10, and 11:20–11:30. By 11:43 IST the same checks were back under 0.3s.

Nothing in the app or the database setting fixes that path. When your two curls match this split, send Railway those times plus the HTTP log line. That is the evidence for the latency on the edge.

Does that check help ?


lazizia
OP

2 months ago

app tls=0.339198s total=0.413537s

noushadalims@Noushadalis-MacBook-Pro ~ % curl -s -o /dev/null -w "google tls=%{time_appconnect}s total=%{time_total}s\n" 'https://www.google.com'

google tls=0.152275s total=0.352581s

noushadalims@Noushadalis-MacBook-Pro ~ %


cptlight

2 months ago

Try the following.

  1. Use a custom domain. Indian ISPs sometimes deprioritize shared Railway domains (*.up.railway.app). Attach a custom domain and CNAME it to your service. This alone resolves most cases.
  2. Disable serverless. If enabled, turn it off. It adds cold-start delays that compound network latency.
  3. Change DNS to 1.1.1.1. Some Indian ISPs have poor DNS paths to Railway's edge. Cloudflare DNS improves initial routing.
  4. Check which edge you hit. Run:

bash

curl -sI 'https://YOUR_APP_DOMAIN/health' | grep -i 'x-railway-edge'

If it returns gru1 (São Paulo) instead of sin1 (Singapore), your traffic is misrouted. This is a known anycast issue with Indian ISPs. Report this to Railway with the header and timings.

  1. Escalate with evidence. Railway's sin1 edge has had intermittent TLS stalls (6–35s) in the handshake while TCP connects fine. When it happens, send Railway:

-The two curl outputs (app TLS vs Google TLS)

-The x-railway-edge header

-Exact IST timestamps

-Your HTTP logs showing the handler finished in ms

  1. Alternative is, put Cloudflare in front. Configure Cloudflare as a proxy (orange cloud) on your custom domain. Indian traffic terminates TLS at Cloudflare's Mumbai POP (~55ms connect, ~131ms TTFB) instead of routing through São Paulo. Cloudflare's backbone carries it to Railway Singapore. This is a potential reliable long-term fix.

Summary:

Custom domain + disable serverless + 1.1.1.1 = immediate mitigation.

x-railway-edge check = diagnosis.

Railway ticket with evidence = potential routing fix.

Cloudflare proxy = permanent workaround.

The problem is edge-side, not your app or DB.


valuemoretec

2 months ago

curl -s -o /dev/null -w "google tls=%{time_appconnect}s total=%{time_total}s\n" 'https://www.google.com'

google tls=0.000000s total=0.000010s

curl -s -o /dev/null -w "app tls=%{time_appconnect}s total=%{time_total}s\n" 'https://www.meeracreations.com/health

app tls=0.000000s total=0.000212s


Status changed to Awaiting User Response Railway • about 2 months ago


valuemoretec

curl -s -o /dev/null -w "google tls=%{time_appconnect}s total=%{time_total}s\n" 'https://www.google.com' google tls=0.000000s total=0.000010s curl -s -o /dev/null -w "app tls=%{time_appconnect}s total=%{time_total}s\n" 'https://www.meeracreations.com/health app tls=0.000000s total=0.000212s

cptlight

2 months ago

Thanks for sharing your results, but those timings don’t look like real network measurements. total=0.000010s is 10 microseconds

  • faster than a local loopback request, let alone a TLS handshake to Google or Railway. That means curl almost certainly didn’t make the network request you intended.

Two things stand out:

Your second command is missing the closing single quote after the URL:

... 'https://www.meeracreations.com/health

Without that ', the shell may have swallowed the rest of the line, so the command may not have executed as expected.

time_appconnect=0.000000 for both Google and your app means the TLS handshake time wasn’t measured at all. It doesn’t mean the handshake was instant.

Could you re-run the exact commands below while the issue is happening, and paste the full output?

curl -v -o /dev/null -w "connect=%{time_connect} appconnect=%{time_appconnect} starttransfer=%{time_starttransfer} total=%{time_total}\n" https://www.google.com

curl -v -o /dev/null -w "connect=%{time_connect} appconnect=%{time_appconnect} starttransfer=%{time_starttransfer} total=%{time_total}\n" https://www.meeracreations.com/health

Also please check which edge you’re hitting:

curl -sI https://www.meeracreations.com/health | grep -i 'x-railway-edge'

If you’re behind Cloudflare, also check:

curl -sI https://www.meeracreations.com/health | grep -i 'cf-ray|server'

The useful comparison is: Google should be fast (TLS under ~0.3s), while your app should show a large time_appconnect (6–35s) when the issue is active. If both are near zero, the test isn’t reaching the network or the problem isn’t occurring at that moment.


Status changed to Awaiting Railway Response Railway • about 2 months ago


2 months ago

This sounds like a routing issue. Could you follow the steps in https://docs.railway.com/networking/troubleshooting/network-diagnostics and send the report here?


Status changed to Awaiting User Response Railway • about 2 months ago


lazizia
OP

2 months ago

Railway Network Diagnostics

Generated : Friday, Oct 2 2026 10:59:06 IST

Endpoint : routing-info-production.up.railway.app


Client IP Info


{

"ip": "223.181.108.228",

"city": "Bengaluru",

"region": "Karnataka",

"country": "IN",

"loc": "12.9719,77.5937",

"org": "AS24560 Bharti Airtel Ltd., Telemedia Services",

"postal": "562114",

"timezone": "Asia/Kolkata",

"readme": "https://ipinfo.io/missingauth"

}


HTTP HEAD request


Status: 200 OK

Status Code: 200

Protocol: HTTP/2.0

Response Headers:

X-Railway-Request-Id: mjYPI-rfSc210AuUmrpb1w

X-Hikari-Trace: sin1.d1nj

X-Railway-Edge: sin1

Content-Type: text/html

Date: Fri, 02 Oct 2026 05:29:07 GMT

Server: railway-hikari


DNS lookup (using system DNS)


; <<>> DiG 9.10.6 <<>> routing-info-production.up.railway.app

;; global options: +cmd

;; Got answer:

;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 22986

;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:

; EDNS: version: 0, flags:; udp: 4096

;; QUESTION SECTION:

;routing-info-production.up.railway.app. IN A

;; ANSWER SECTION:

routing-info-production.up.railway.app. 60 IN A 69.46.46.119

;; Query time: 60 msec

;; SERVER: fe80::1%14#53(fe80::1%14)

;; WHEN: Fri Oct 02 10:59:07 IST 2026

;; MSG SIZE rcvd: 83


DNS lookup (using Cloudflare)


; <<>> DiG 9.10.6 <<>> @1.1.1.1 routing-info-production.up.railway.app

; (1 server found)

;; global options: +cmd

;; Got answer:

;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 44248

;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:

; EDNS: version: 0, flags:; udp: 1232

;; QUESTION SECTION:

;routing-info-production.up.railway.app. IN A

;; ANSWER SECTION:

routing-info-production.up.railway.app. 60 IN A 69.46.46.119

;; Query time: 52 msec

;; SERVER: 1.1.1.1#53(1.1.1.1)

;; WHEN: Fri Oct 02 10:59:07 IST 2026

;; MSG SIZE rcvd: 83


Traceroute


1 dsldevice.lan (192.168.1.1) 9.560 ms 2.987 ms

2 10.240.15.251 (10.240.15.251) 44.578 ms 7.093 ms

3 * *

4 dsl-kk-static-077.206.95.61.airtelbroadband.in (61.95.206.77) 6.431 ms 5.841 ms

5 116.119.119.205 (116.119.119.205) 77.473 ms 44.034 ms

6 60068.sgw.equinix.com (27.111.230.50) 45.125 ms 46.128 ms

7 vl221.sgp-m1-dist-2.cdn77.com (79.127.194.188) 44.430 ms *

8 69.46.46.119 (69.46.46.119) 44.919 ms 45.934 ms


Ping (n=10)


PING routing-info-production.up.railway.app (69.46.46.119): 56 data bytes

64 bytes from 69.46.46.119: icmp_seq=0 ttl=58 time=46.699 ms

64 bytes from 69.46.46.119: icmp_seq=1 ttl=58 time=45.437 ms

64 bytes from 69.46.46.119: icmp_seq=2 ttl=58 time=63.613 ms

64 bytes from 69.46.46.119: icmp_seq=3 ttl=58 time=42.251 ms

64 bytes from 69.46.46.119: icmp_seq=4 ttl=58 time=45.676 ms

64 bytes from 69.46.46.119: icmp_seq=5 ttl=58 time=47.417 ms

64 bytes from 69.46.46.119: icmp_seq=6 ttl=58 time=41.723 ms

64 bytes from 69.46.46.119: icmp_seq=7 ttl=58 time=45.115 ms

64 bytes from 69.46.46.119: icmp_seq=8 ttl=58 time=43.752 ms

64 bytes from 69.46.46.119: icmp_seq=9 ttl=58 time=44.060 ms

--- routing-info-production.up.railway.app ping statistics ---

10 packets transmitted, 10 packets received, 0.0% packet loss

round-trip min/avg/max/stddev = 41.723/46.574/63.613/5.931 ms

Completed


Status changed to Awaiting Railway Response Railway • about 2 months ago


sam-a
EMPLOYEE

2 months ago

Thanks for the report. It shows your connection reaching our Singapore edge directly over Airtel, with around 46 ms round-trip and no packet loss, which is the expected routing from Bengaluru.

That capture looks like it was taken during a normal period, so it doesn't show the slowdown itself. Please run the diagnostics tool again while the slowness is actually happening and post that report here, along with the time it occurred in IST.


Status changed to Awaiting User Response Railway • about 2 months ago


Railway
BOT

a month ago

This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!

Status changed to Solved Railway • about 1 month ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...