platform · inferred from evidence
GCP
Google Cloud Platform, a suite of cloud computing services.
- Flipt 1.9.0 Postgres dial tcp lookup postgres device or resource busy
This was a Flipt regression fixed in v1.9.1 and v1.10.1, upgrade to either of those patch releases and the Postgres connection works again. The reporter confirmed v1.9.1 fixed it in both local Rancher Desktop and GCP clusters. If you must s
- Cloud SQL Proxy: "couldn't connect ... dial tcp ...:3307: getsockopt: connection timed out"
The proxy intentionally dials the Cloud SQL instance on port 3307 (local clients still use 3306; 3307 isolates proxy traffic on the instance side), so your firewall/VPC must allow egress from the proxy host to the instance IP on port 3307.
- ChromaDB DuplicateIDError on memgpt load (duplicate embedding IDs)
This was a Flipt regression fixed in v1.9.1 and v1.10.1, upgrade to either of those patch releases and the Postgres connection works again. The reporter confirmed v1.9.1 fixed it in both local Rancher Desktop and GCP clusters. If you must s
- Google Secret Manager Could not load the default credentials from AWS Lambda nodejs client
The Google-recommended approach for AWS-to-GCP auth is Workload Identity Federation instead of long-lived service-account keys: configure a workload identity pool for the AWS account and use google.auth with the external-account credentials
- `google-cloud-documentai` imports fail in GCP notebooks: kernel Python does not match pip
The package is fine — the environment is the problem. If you selected a GCP notebook instance pre-loaded for Apache Beam, its Jupyter kernel runs a different Python (`/root/apache-beam-custom/bin/python`) than the one pip installed into (`/
- Crawlee SQLite "disk I/O error" in container (Cloud Run)
SQLite's WAL (write-ahead logging) mode does not work on Cloud Run's filesystem. Disable it with the environment variable `APIFY_LOCAL_STORAGE_ENABLE_WAL_MODE=false` before starting the crawler. The reporter confirmed Apify then runs correc