fly deploy issues and slow app performance

I’m attempting to deploy my .net core API app using fly deploy and fly deploy --remote-only but it’s unable to complete the upload of layers of my docker container. Most layers seem to push successfully, but it seems to be blocked on pushing one layer which prevents the upload from completing. I have waited 500 seconds before cancelling the upload attempt.

When running a traceroute, this is what I get:

traceroute passm8-prod-backend.fly.dev
traceroute to passm8-prod-backend.fly.dev (66.241.124.48), 64 hops max, 40 byte packets
 1  rt-ax82u-7100 (192.168.50.1)  5.245 ms  4.813 ms  4.640 ms
 2  100.101.128.1 (100.101.128.1)  13.805 ms  13.495 ms  12.636 ms
 3  cgnat.20wha.qld.aussiebb.net (202.142.143.119)  12.271 ms  12.166 ms  12.745 ms
 4  be50-3999.cfl3.20wha.qld.aussiebb.net (119.18.32.104)  12.971 ms  12.456 ms  12.345 ms
 5  10.241.13.167 (10.241.13.167)  23.819 ms  22.864 ms  23.585 ms
 6  * * *
 7  * * *
 8  * * *
 9  * * *
10  * * *
11  * * *
12  * * *
13  * * *
14  * * *
15  * * *
16  * * *
17  * * *
18  * * *
19  * * *
20  * * *
21  * * *
22  * * *
23  * * *
24  * * *
25  * * *
26  * * *
27  * * *
28  * * *
29  * * *
30  * * *
31  * * *

I did manage to upload successfully last night, but now my app’s performance seems to have degraded on SOME but not all endpoints. When running locally I’m able to get response times in about 200ms, but when running remotely it takes about 5-10 seconds. I’m unsure if this is a related issue or not, but thought it worth mentioning.

I am not sure what other information is necessary to provide at this point. It seems to be something that should be taken to the fly team directly, but I have to pay for this ability - seems strange given that it seems to be an infrastructure related issue rather than seeking support on how to do something.

1 Like

I wonder if flyctl deploy --local-only would help.

How big is your Docker image?

Hi halfer,

My docker image is 155Mb, so not huge. I have since raised a support request with fly.io and they moved my builder from syd to sin and this has resolved the issue. It now takes seconds to push the docker layers rather than timing out after 500 seconds. I am curious what the problem was because it was actually stating a 500 server error stating something about the graphql endpoint timing out (sorry I didn’t capture the actual message).

1 Like

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.