Problem
During the E2E run on 2026-03-12, DigitalOcean reported a droplet limit exceeded error when trying to provision claude and opencode in parallel with other agents:
Fatal: DigitalOcean API error 422 for POST /droplets: {"id":"unprocessable_entity","message":"creating this/these droplet(s) will exceed your droplet limit"}
The suite runs agents in parallel batches of 5. Batch 1 starts 5 agents simultaneously for each cloud. While some droplets are already running (zeroclaw, openclaw, codex, kilocode, hermes), a new batch spins up more — hitting the account's droplet limit.
Affected
- digitalocean/claude — FAIL (batch 1, hit limit immediately)
- digitalocean/opencode — FAIL (batch 2, limit still at capacity)
Possible fixes
- Increase the DigitalOcean droplet limit via the DO dashboard (Settings → Limits → Request increase)
- Add retry logic in
sh/e2e/lib/provision.sh for 422 errors — wait for previous batch to tear down before retrying
E2E Results Context
- aws: SKIP (no credentials)
- hetzner: 8/8 PASS
- digitalocean: 6/8 PASS (claude FAIL, opencode FAIL — both due to droplet limit)
- gcp: 8/8 PASS
- sprite: 7/8 PASS (opencode FAIL — separate issue, binary not found)
- Total: 29/32 PASS
-- qa/e2e-tester
Problem
During the E2E run on 2026-03-12, DigitalOcean reported a droplet limit exceeded error when trying to provision
claudeandopencodein parallel with other agents:The suite runs agents in parallel batches of 5. Batch 1 starts 5 agents simultaneously for each cloud. While some droplets are already running (zeroclaw, openclaw, codex, kilocode, hermes), a new batch spins up more — hitting the account's droplet limit.
Affected
Possible fixes
sh/e2e/lib/provision.shfor 422 errors — wait for previous batch to tear down before retryingE2E Results Context
-- qa/e2e-tester