Terraform "timeout while waiting for plugin to start" in CI

Terraform launches each provider as a subprocess and waits for it to report ready. If the handshake does not complete in time -- a slow or starved runner, a partial download -- it errors.

What this error means

init/plan fails with "timeout while waiting for plugin to start" or "Failed to instantiate provider". It can be intermittent, passing on retry, which points at a transient launch problem rather than bad config.

terraform

Error: Failed to instantiate provider
"registry.terraform.io/hashicorp/aws" to obtain schema: timeout while waiting
for plugin to start

Common causes

Resource-starved or slow runner

A CPU/memory-starved runner, or heavy parallelism, can delay the plugin handshake past the startup timeout.

Partially downloaded or incompatible binary

A truncated provider download or an arch mismatch (e.g. arm64 vs amd64) leaves a binary that cannot start.

How to fix it

Retry and reduce parallelism

Re-run init/plan and lower -parallelism so plugin startup is not starved; clear the cache if a binary is suspect.

Terminal

rm -rf .terraform/providers
terraform init -input=false
terraform plan -input=false -parallelism=5

Ensure the correct provider binary

Confirm the provider supports the runner OS/arch.
Clear .terraform/providers and re-init to redownload cleanly.
Avoid extreme parallelism on small runners.

How to prevent it

Use adequately sized runners for large provider graphs.
Keep provider downloads clean (cache validation, correct arch).
Run plan/apply on a platform that auto-retries transient plugin-start timeouts.