Fresh tester public gateway URL stays 502 until the gateway node's route to the new VM propagates #272
Labels
No labels
meeting-notes
meeting-transcript
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
lhumina_code/home#272
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
After provisioning or re-provisioning a tester, the public URL (name.gent01.qa.grid.tf) can return 502 for a long time (seen 35 plus minutes) even though the tester is fully installed and its login gate works. Verified: the tester backend serves 302 over mycelium from other nodes, DNS resolves to the gateway node, and the on-chain name contract is clean (a single live contract, no orphans). The gateway node simply cannot reach the new VM yet. The gateway proxies per request, so re-deploying the gateway does not help; it is purely the gateway-node to fresh-VM mycelium route propagating. We should confirm whether this reachability can be sped up or pre-warmed (a ready VM pool would sidestep it). Separately, redoing a gateway for an existing user is fragile today: it needs the old name contract cancelled and the old gateway deleted by hand, compounded by the nonce issue (lhumina_code/hero_compute#132) and the empty-fqdn issue (lhumina_code/hero_compute#133).