Open GeraldLoeffler opened 3 weeks ago
It looks this error is caused by either the nitric/awstf
provider failing to start or the two processes failing to communicate. We may want to add more output to the provider/CLI to show any other details about why it failed.
We've also seen other instances of this issue: https://discord.com/channels/955259353043173427/1285193296850845800/1286319877736628356
@GeraldLoeffler is this failure consistent or intermittent? Its possible there may be a race condition between when we start the provider on your machine and when our CLI tries to connect.
The context deadline for provider startup was 5 seconds. Given we've seen this issue be intermittent in the past, there is a chance the timeout was too short. Any other issue in the provider, such as a panic, is likely to have printed to the console, so slow startup seems like a likely cause.
I've extended the timeout in the Podman support PR https://github.com/nitrictech/cli/pull/815 so we can see if that helps.
@GeraldLoeffler would you mind testing this again with v1.55.2 of the CLI?
Feature Request
Suggestion
Support development with the
nitric
CLI on machines with a network configuration that requires usage of an explicit web proxy. Typically, such a proxy is configured with environment variablesHTTP_PROXY
,HTTPS_PROXY
andno_proxy
.Currently, with Docker Desktop on Apple Silicon in such an environment,
nitric up
fails witherror failed to connect to provider: context deadline exceeded
Value
In corporate setups, explicit web proxies are unfortunately not uncommon. This requires correct configuration of all client machines, including development machines, and all web browsers. Nitric development should just work on a development machine configured like that.
Docker Desktop in itself works without issues in the presence of explicit web proxies.
Alternatives
Podman instead of Docker Desktop might work in such a setting?
Other info
Complete output of
nitric up
:Versions used: